Home Economics American Trypanosomiasis Chagas Disease, Second Edition: One Hundred Years of Research
Recombination in natural populations
Inter-lineage (inter DTU) recombination: TcV and TcVI
The clonal theory of parasitic protozoa implied that genetic exchange was either absent or a rare event in T. cruzi and of little or no epidemiological significance.1,2 This hypothesis was supported by several features of the genetic diversity of T. cruzi, including strong linkage disequilibrium, identical genotypes spread over vast geographical distances, and phylogenetic correlation between independent sets of genetic markers. Nevertheless, early MLEE studies of isolates from Bolivia and Paraguay, now incorporated within lineage TcV and TcVI, respectively, revealed highly distinctive, heterozygous MLEE profiles, with at least one corresponding homozygous profile seen among other isolates from the same locality.13,27 For example, the dimeric enzyme glucose phosphate isomerase (GPI) had the profile of three equidistantly separated bands, with a central band that was more intense as would be expected for recombinant strains. Furthermore, this profile was sustained in clones so could not be due to a mixture of populations. The TcV and TcVI profiles were similar in that both showed a high level of heterozygosity but they also had some distinguishing isoenzyme bands.
TcV and TcVI have an unusual geographical distribution: they were found to be particularly common in the foothills of the Andes and the greater Gran Chaco regions of Bolivia, Chile, Paraguay, northern Argentina, and in the extreme south of Brazil, where there were wider fluctuations in environmental temperature than in the Amazon and Atlantic forests. This suggested that the heterozygous MLEE profiles might be adaptive, giving enhanced fitness in triatomine bugs through metabolic flexibility over a range of environmental temperatures. Such an adaptive advantage was proven to occur in other systems, e.g., in fish moving between warm and cold climatic conditions. Hybridization is also known to be associated with the generation of
new phenotypes in Leishmania sp.28 The metabolic significance of the predominantly heterozygous profiles of TcV and TcVI was explored by purifying the GPI isoenzymes and testing their catalytic rates at different temperatures. Although the isoenzymes certainly differed in temperature stability, as was readily seen directly by their differing persistence on incubated MLEE gels, there was no proof of differences in catalytic efficiency at diverse temperatures for the purified isoenzymes.29 Nevertheless, the maintenance of high levels of heterozygosity in TcV and TcVI suggests that such experiments to test for associations with fitness remain worthwhile.
The hybrid nature of TcV and TcVI was confirmed by other molecular methods, notably by sequencing of housekeeping genes.30-32 Comparison of nucleotide sequences showed that TcV and TcVI must have arisen through hybridization between genetically distinct parents because they were found to possess fully intact alleles from two other DTUs (TcII and TcIII) that are so distinct that they could not possibly have arisen independently. As in the experimental hybrids described in the previous section, sequencing of kDNA maxicircle genes showed they were inherited uniparentally by TcV and TcVI, with the TcIII parent being the donor in each case.
Lewis et al.32 conducted a comprehensive analysis of TcV and TcVI based on nuclear and mitochondrial gene sequences as well as a panel of 28 microsatellite loci. These markers evolve at very different rates, which allowed the relationships between hybrid and nonhybrid DTUs to be studied with high resolution at multiple timescales. Overall, TcV and TcVI each closely resembled the sterile F1 meiotic progeny of a relatively recent cross between TcII and TcIII. The hybrid genotypes were found to be highly heterozygous, with low intra-DTU diversity. Very few unique polymorphisms were identified, consistent with limited clonal diversification after hybrid formation. Consistent with this, multilocus sequence analysis of coding genes showed that intra-TcV and intra-TcVI differences were essentially restricted to loss of heterozygosity (LOH) events, such that some strains had become homozygous for TcII or TcIII parental alleles.33
The close genetic similarity between TcV and TcVI has made it difficult to determine whether they are the products of a single hybridization event followed by limited clonal diversification or of two independent events involving genetically similar TcII and TcIII parental strains. Analyses of housekeeping gene sequences supported the hypothesis of a single hybridization event followed by divergence to form TcV and TcVI.34,35 In contrast, microsatellite data based on five loci36 and other coding gene phylogenies37 have indicated that two independent events are the most likely explanation. However, these studies used relatively few strains and/or low resolution data such that they were probably underpowered to address the question. The multilocus microsatellite analysis conducted by Lewis et al.32 showed that the two hybrid groups had highly distinct profiles composed of different combinations of alleles that were shared with TcII and TcIII strains. This, combined with the many known coding sequence differences, strongly suggested that TcV and TcVI had not evolved by diversification from a common ancestor. Two scenarios for the origins of TcV and TcVI were proposed to fit these data: two independent hybridization events between distinct TcII and TcIII strains or two independent progeny from a single TcII X TcIII cross.
Comparison of hybrid and parental genotypes has been used as a way to identify extant TcII and TcIII strains that most closely resemble the ancestral parents that hybridized to form TcV and TcVI. For example, analysis of 5S rDNA sequences has shown that the TcII-like allele for these sequences found in the hybrid strains is more similar to TcII isolates from Bolivia and Chile than to others from Brazil.38 Similarly, TcII-like GPI sequences and microsatellite alleles in the hybrids were phylogenetically closer to those in modern strains from Chile, Paraguay, and Bolivia than those from Brazil.32 The TcIII alleles present in the hybrids supported a closer genetic relationship with modern TcIII from Bolivia, Paraguay, or Peru than strains from Brazil, Colombia, or Venezuela. Mitochondrial kDNA maxicircle sequences, inherited by the hybrids uniparentally from TcIII, point more specifically to a smaller subset of TcIII strains, most commonly found in Paraguay.31,32 Although the current distribution of TcII and TcIII strains may not coincide with the situation when TcV/VI were formed, their genotypes are most consistent with an origin in the Gran Chaco or adjoining Andean valleys in southwestern South America. Emerging data show that the rare occurrence of TcV and TcVI genotypes in Colombia is almost certainly a result of dispersal rather than more ancient (or additional) hybridization events (L. Messenger, unpublished data). This does, however, demonstrate that the discovery of new TcV or TcVI strains, particularly from sylvatic sources, will necessitate reevaluation of their evolutionary origins.
Intuitively, the more recent the natural hybridization event(s) that gave rise to TcV and TcVI, the more likely it is that new natural hybrid lineages could be expected to emerge. Moreover, given the abundance of hybrids, particularly TcV, in human infections across large endemic areas, there may be a significant risk associated with newly emergent recombinant strains. In an effort to estimate the evolutionary timeframe for the origin of TcV and TcVI, Lewis et al.32 conducted a phylogenetic molecular clock analysis on a large set of nuclear and mitochondrial sequences from hybrid and non-hybrid strains. These authors concluded that the extant TcV and TcVI strains last shared a common ancestor within approximately the last 60,000 years. Such a recent timescale is close to the limits of resolution for the sequences that were used, however, so the actual origin may well be more recent, especially considering the lack of diversity at microsatellite loci, which typically have very fast mutation rates. Furthermore, the calibration point for the molecular clock used may have been overly conservative.39 Such a recent origin for natural TcII—III hybrids, together with evidence for genetic exchange within TcI in the laboratory17 and the field40,41 indicate that sexual recombination continues to be available as a reproductive mode to a broad range of T. cruzi strains.
The ecological circumstances of the origin of TcV and TcVI are not known and can only be inferred based on the balance of probabilities. These hybrids are strongly associated with domestic vectors, humans, and domestic animals. Records of nondomestic TcV/VI are extremely rare and wild populations of their principal vector, T. infestans, appear to be overwhelmingly associated with TcI.42 The formative hybridization events must have occurred in either a mammalian host or a triatomine bug coinfected with TcII and TcIII. However, wild populations of these DTUs have apparently little overlap.6 Lewis et al.32 suggested that the most likely circumstances for TcII—TcIII coinfections to occur were in domestic transmission cycles because they provide an environment where they can and do overlap. Even though molecular clock analysis indicated the hybrids arose prior to the arrival of humans in South America there is enough leeway in the dating estimates (see above) to accommodate a more recent, anthropogenic origin. The apparent lack of modern day genetic diversity within both hybrid DTUs from isolates covering a vast geographic area is most plausibly a result of a recent spread of TcV and TcVI as two clonal lineages in association with the spread of domiciliated T. infestans, which evidence indicates is itself due to human activities and population movements.43
If TcV and TcVI were, like the experimental hybrids, products of genome fusion of diploids to yield aneuploid progeny, the redundant extra copies of genes might confer versatility and evolutionary advantage, in that those genes would potentially be free to evolve rapidly and independently and to acquire alternative independent functions. Accordingly, flow cytometric analysis of DNA content and multilocus genotyping has been applied to natural isolates of T. cruzi representing the known T. cruzi lineages.18 Unlike experimental TcI hybrids, the natural TcV and TcVI hybrids were found to have DNA contents consistent with diploidy and equivalent to the average DNA contents of isolates representative of their TcII and TcIII parents. All the available evidence indicates that TcV and TcVI are typical (sterile) F1 meiotic progeny — they possess one TcII allele and one TcIII allele for most loci and the limited cases of homozygosity are best explained by LOH through gene conversion.12,32 There are, therefore, fundamental differences between the naturally occurring hybrid strains and the experimental hybrids both in terms of allelic inheritance patterns and overall DNA content. It is not clear whether these differences reflect the operation of distinct mechanisms of genetic exchange. It is possible that developmental cues that would cause the experimental hybrids to return to diploidy were absent under laboratory conditions. Alternatively, if TcV and TcVI are indeed derived from fusion of diploids, genome reduction may have progressed sufficiently to result in reversion to a heterozygous diploid state.32 The mechanism of genetic exchange in natural populations is thus yet to be fully understood and potentially differs from that documented so far from experimental observations.
While the precise biological circumstances that governed the origin and spread of TcV and TcVI remain to be fully understood, the epidemiological impact of these hybrids is striking. They are highly abundant in domestic transmission cycles and TcV predominates among human infections in Bolivia and northern Argentina. More limited data indicate these hybrids are also common causes of human infection in Chile, Paraguay, and southern Brazil. Chagas disease manifestations are often severe in these regions and chagasic cardiomyopathy, megaoesophagus, megacolon, and congenital transmission are all common. In this context, the epidemiological importance of genetic exchange in T. cruzi is clear: it has been and may still be profoundly important.
|< Prev||CONTENTS||Next >|