Molecular Biology and Evolution 18:2250-2259 (2001)
© 2001 Society for Molecular Biology and Evolution
Maximum-Likelihood Divergence Date Estimates Based on rRNA Gene Sequences Suggest Two Scenarios of Trypanosoma cruzi Intraspecific Evolution
Departamento de Microbiologia, Imunologia e Parasitologia, Escola Paulista de Medicina, Universidade Federal de São Paulo;
Departamento de Medicina Tropical, Instituto Oswaldo Cruz, Fundação Oswaldo Cruz, Rio de Janeiro;
Departamento de Bioquímica, Instituto de Química, Universidade de São Paulo
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
The phylogenetic relationships of Trypanosoma cruzi strains were inferred using maximum-likelihood from complete 18S rDNA sequences and D7-24S
rDNA regions from 20 representative strains of T. cruzi. For this we sequenced the 18S rDNA of 14 strains and the D7-24S rDNA of four strains and aligned them to previously published sequences. Phylogenies inferred from these data sets identified four groups, named Riboclades 1, 2, 3, and 4, and a basal dichotomy that separated Riboclade 1 from Riboclades 2, 3, and 4. Substitution models and other parameters were optimized by hierarchical likelihood tests, and our analysis of the 18S rDNA molecular clock by the likelihood ratio test suggests that a taxa subset encompassing all 2,150 positions in the alignment supports rate constancy among lineages. The present analysis supports the notion that divergence dates of T. cruzi Riboclades can be estimated from 18S rDNA sequences and therefore, we present alternative evolutionary scenarios based on two different views of T. cruzi intraspecific divergence. The first assumes a faster evolutionary rate, which suggests that the divergence between T. cruzi I and II and the extant strains occurred in the Tertiary period (37–18 MYA). The other, which supports the hypothesis that the divergence between T. cruzi I and II occurred in the Cretaceous period (144–65 MYA) and the divergence of the extant strains occurred in the Tertiary period of the Cenozoic era (65–1.8 MYA), is consistent with our previously proposed hypothesis of divergence by geographical isolation and mammalian host coevolution. |
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Chagas' disease, for which there is no vaccine or cure at the chronic stage, affects 18 million people in the Americas and 30 million are still at risk of infection by its causative agent, the protozoan Trypanosoma cruzi. It is therefore essential to establish whether the different strains of T. cruzi can be typed to identify groups of strains that can be associated with the different epidemiological characteristics and clinical manifestations, namely, cardiopathy, megacolon, megaesophagous, and indeterminate forms (Wendel et al. 1992
). These might be relevant in the future for predictive purposes of the disease outcome and for differential treatment. Initial attempts at typing and grouping of the strains were done using isoenzyme profiles (Miles et al. 1978
) and restriction fragment length patterns of the mitochondrial or kinetoplast DNA (kDNA) (Morel et al. 1980
). Typing by isoenzymes led to the establishment of three groups called zymodemes Z1, Z2, and Z3 (Miles et al. 1978
). Comparison of nuclear genes suggested that two major groups can be identified, named T. cruzi I (sylvatic cycle) and T. cruzi II (domestic cycle) (Satellite-Meeting 1999
), corresponding to lineages 2 and 1, respectively, determined by rDNA (Souto et al. 1996
; Briones et al. 1999
). Recently, typing by microsatellite and by low-stringency single-specific primer-polymerase chain reaction (LSSP-PCR) from kDNA segments helped to determine that chagasic patients may be infected with multiclonal strains and that in the same mammalian host different strains may infect different organs (Gomes et al. 1998
; Vago et al. 2000
). In a cohort of 22 patients with monoclonal infections and digestive and cardiac symptoms, it was determined that infective strains belonged to T. cruzi II (lineage 1 from Souto et al. 1996
; Briones et al. 1999
) (E. Lages-Silva, E. Chiari, and A. M. Macedo, personal communication). In this context, it is relevant to determine whether the subdivision T. cruzi I and T. cruzi II corresponds to different phylogenetic groups or clades using explicit models of phylogenetic evolution as implemented in maximum-likelihood analysis (Felsenstein 1981
; Posada and Crandall 1998
). If this subdivision is as deep as suggested by previous studies (Briones et al. 1999
), slowly evolving markers, such as rDNA sequences, should be employed (Hillis, Mable, and Moritz 1996
).
Here we studied the evolution of T. cruzi strains, based on maximum-likelihood analysis of 18S rDNA sequences from 20 strains of T. cruzi, and identified four subgroups which we named as Riboclades. Also, we expanded the analysis of the highly variable region, D7, within the 24S rDNA (D7-24S) by sequencing the corresponding segments from four strains of the zymodeme 3 subgroup (Miles and Cibulskis 1986
; Mendonça et al. 2001
). Maximum-likelihood parameters and the molecular clock of corresponding phylogenies were optimized and tested by likelihood ratio tests. Our analysis suggests that divergence times can be estimated from particular 18S rDNA subsets. Given the external calibration of the rDNA clock from metazoan 18S rDNA, the divergence times estimated here support our proposed hypothesis of geographical isolation of the two major groups of strains in the last 84 Myr and their possible coevolution with the mammalian fauna of the Americas during the Cenozoic (Briones et al. 1999
). If the 18S rRNA gene in T. cruzi is assumed to evolve twofold to fourfold faster than in other organisms, a scenario consisting of a more recent divergence (18–16 MYA) is favored (Machado and Ayala 2001)
.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
T. cruzi Strain Typing by D7-24S rDNA Amplification
Parasites were cultured and total DNA extracted as described (Souto and Zingales 1993
). Typing was done by PCR amplification of the variable region, D7, with primers D71 5'-AAGGTGCGTCGACAGTGTGG-3' and D72 5'-TTTTCAGAATGGCCGAACAGT-3' as described. Amplified products were analyzed in 7.5% polyacrylamide gel electrophoresis and UV visualization after ethidium staining (Souto and Zingales 1993
).
Sequencing of 18S rDNA and D7-24S rDNA
The 18S rDNA was amplified using universal eukaryotic primers and cloned into pUC18 using the Sureclone kit (Amersham-Pharmacia). D7-24S rDNA sequences were amplified using specific primers as described earlier. Templates were sequenced in a ABI377/96 automated sequencer (Applied Biosystems) using fluorescent BigDye dideoxy-chain terminators. Chromatograms, derived from complete sequences of at least three clones from each strain, were used for contiguous assembly and finishing of high quality sequences (Phred scores above 30, which correspond to less than 1 estimated error per 1,000 bases sequenced) using Phred + Phrap + Consed (Gordon, Abajian, and Green 1998
). Chromatograms are available upon request by e-mail (marcelo@ecb.epm.br). Sequences determined in this study were deposited in GenBank as follows: 18S rDNA sequences: T. cruzi MT3869, AF303660; T. cruzi Silvio X10 cl1, AF303659; T. cruzi Y, AF301912; T. cruzi YuYu, AF245380; T. cruzi MT4166, AF292942; T. cruzi MT4167, AF288661; T. cruzi MT3663, AF288660; T. cruzi NRcl3, AF228685; T. cruzi CL Brener, AF245383; T. cruzi Dm28c, AF245382; T. cruzi CA-1, AF245381; T. cruzi G, AF239981; T. cruzi Colombiana, AF239980; and T. cruzi Sc43cl1, AF232214. The D7-24S rDNA sequences here determined were: T. cruzi 4167, AF288665; T. cruzi 4166, AF288664; T. cruzi 3869, AF288663; and T. cruzi 3663, AF288662. Details on the host and geographic origin of the strains can be found elsewhere (Souto et al. 1996
).
Phylogenetic Inference
Alignments were made manually using the sequences determined in this study and other sequences downloaded from GenBank: T. cruzi Peru X53917 (Fernandes, Nelson, and Beverley 1993
); T. cruzi Silvio X10, AJ009147; T. cruzi Vinch89, AJ009149; T. cruzi CANIII AJ009148; T. cruzi marinkellei, AJ009150 (Stevens, Noyes, and Gibson 1998
); and T. rangeli San Agustin, AF065157 (Briones et al. 1999
). Sequences were aligned by eye using seaview sequence editor for UNIX (Galtier, Gouy, and Gautier 1996
). Alignments are available upon request by e-mail (marcelo@ecb.epm.br). Phylogenies were inferred using maximum-likelihood estimates using PAUP 4.0 (Swofford 1998
). Maximum-likelihood parameters, such as model, base frequencies, shape of gamma distribution, and proportion of invariant sites, were optimized using the hierarchical likelihood test with Bonferroni correction implemented in Modeltest 3.0 (Posada and Crandall 1998
). Bootstrap analysis to test for monophyly was done using PAUP 4.0 with removal of invariant and gapped positions, and parameters were estimated for each bootstrap replicate. Likelihood ratio tests for the molecular clock were performed as described (Felsenstein 1988
; Swofford et al. 1996
) and PAUP 4.0 was used to obtain the likelihood scores of each phylogeny. Confidence intervals for divergence time estimates were obtained using QDate 1.11 (Rambaut and Bromham 1998
), where the divergence time estimates of quartet pairs in the input file were calculated using percent divergence from fossil calibrations as described by Escalante and Ayala (1995)
. QDate input run parameters were obtained by Modeltest analysis.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Maximum Likelihood Inference of 18S rDNA and D7-24S rDNA Phylogenies
The 18S rDNA alignment contained sequences of 20 strains of T. cruzi, from which 14 were determined in the present study and six were downloaded from GenBank (see Materials and Methods). Trypanosoma rangeli was used as an outgroup for rooting the T. cruzi ingroup. The model fit estimate was done using Modeltest 3.0 and PAUP 4.0. The model selected for this data set was the TrN (Tamura and Nei 1993
) which uses the following parameters: equal frequencies of the nucleotides, the rate matrix (A–C = 1, A–G = 0.8591, A–T = 1, C–G = 1, C–T = 4.9784, G–T = 1), the proportion of invariable sites = 0.7951, and the gamma distribution (Yang 1996
) shape parameter () = 0.6325. The resulting phylogeny (LnL = -3,958.85482) (fig. 1A
) was searched by a heuristic maximum-likelihood tree-bisection–reconnection swapping algorithm (PAUP 4.0). The D7-24S rDNA phylogeny (LnL = -390.16531) (fig. 1B
) was obtained by the same approach with equal base frequencies, transition rate = 1.4056, proportion of invariable sites = 0, and equal rates for all sites for the K81 selected as the model (Kimura 1981
). For the bootstrap analysis, we considered that each bootstrap replicate contained a resampling of the original data set. Therefore, the parameters optimized for the original data set cannot be used to infer the 100 different trees for each bootstrap replicate data set. For this, the parameters would have to be estimated for each data set, which takes an impractical computational time. Therefore, we excluded T. rangeli and used T. cruzi marinkellei as an outgroup, given that T. cruzi marinkellei is a trypanosome of bats clearly distinguished from T. cruzi lineages (Brisse et al. 2000b
), which is also suggested by the topology of the tree in figure 1A.
Also, we excluded the gapped and invariant sites. The model, F81 (Felsenstein 1981), was selected from a Modeltest analysis where rates are equal for every substitution type and site nucleotide frequencies fA = 0.1787, fC = 0.3320, fG = 0.1790, fT = 0.3130, and proportion of invariant sites = 0. Using this computationally faster approach, the parameters that have to be estimated for each bootstrap replicate are the nucleotide frequencies. The results are included in the 18S rDNA phylogeny in figure 1A
and they suggest that four groups, named here as Riboclades 1, 2, 3, and 4, can be identified. According to this analysis, the best results were obtained for Riboclade 1, with bootstrap supports of 98% and 100% in the 18S rDNA and D7-24S rDNA phylogenies, respectively; and for Riboclade 3, with 96% and 71% in 18S rDNA and D7-24S rDNA phylogenies, respectively.
|
Phylogenies shown in figure 1 suggest that T. cruzi has a basal dichotomy which separates Riboclade 1 from Riboclades 2, 3, and 4, with good bootstrap support. The Riboclades of our study were then compared with data obtained with a previously described typing methodology for T. cruzi strains (Souto et al. 1996
) to establish a correspondence among subgroups. Typing of T. cruzi strains based on band sizes of the amplified D7-24S rDNA (fig. 2
) reveals four band patterns that correspond to the Riboclades described in the phylogenies shown in figure 1
.
|
Likelihood Ratio Tests For Molecular Clock
In our previous study we proposed a hypothesis for the evolution of the two main T. cruzi groups based on molecular divergence time estimates (Briones et al. 1999
). According to this hypothesis, the differential distribution of lineages 1 (T. cruzi II) and 2 (T. cruzi I) in placental and marsupial hosts, noticed in North America (Clark and Pung 1994
) and in Brazil (Fernandes et al. 1999
) could be correlated with the evolution of the mammalian fauna in the Americas during the Cenozoic (Briones et al. 1999
). Here, we tried to improve our divergence time estimates by analyzing the 18S rDNA sequences of strains of T. cruzi, representative of different groups using the likelihood ratio test for the molecular clock. For this we partitioned the original 18S rDNA into several different subsets (table 1 ). The models selected in the subset analysis were as in table 1 : TrN (Tamura and Nei 1993
), JC (Jukes and Cantor 1969
), K80 (Kimura 1980
), HKY (Hasegawa, Kishino, and Yano 1985
), F81 (Felsenstein 1981
), and K81 (Kimura 1981
). The partitioning was first made by dividing the gene into four larger parts and then into more specific positions, such as informative and ungapped positions only. The model fit for each partition was done using Modeltest (Posada and Crandall 1998
) and the nonclocklike and clocklike trees were inferred for the individual subsets. Our analysis shows that for the complete 18S rDNA data set (subset 1 in table 1
), as used to infer the 18S rDNA tree in figure 1A,
the clocklike tree is rejected. The clocklike trees are rejected in subsets 4, 6, 7, 8, 9, and 11. However, clocklike trees are not rejected in subsets 2, 3, 5, 10, 12, 13, and 14. This suggests that exclusion of positions 1001–1500 of the 18S rDNA and that replacement of T. rangeli by T. cruzi marinkellei as rooting outgroup produces clocklike trees that cannot be rejected with a 95% confidence interval. This means that the increase in the likelihood score, obtained when branch lengths are unconstrained by ultrametric tree properties, does not significantly improve the likelihood of the estimate. Interestingly, in data set 12 (table 1
), when sequences T. cruzi La Cruz, X10; Vinch89; CANIII (Stevens, Noyes, and Gibson 1998
) and T. rangeli are removed (and trees rooted by T. cruzi marinkellei), a clocklike tree is obtained, which includes all 2,150 positions that cannot be rejected (table 1
).
|
Divergence Time Estimates
On the basis of our likelihood ratio tests for the molecular clock, we estimated the divergence times from the subset 12 clocklike tree (table 1 ). This subset was chosen because the available evolutionary rate of the 18S rDNA reflects the evolution of the gene as a whole and not simply the highly divergent regions (Escalante and Ayala 1995
) and because exclusion of taxa is preferable to exclusion of sites. The distances were obtained from the phylogeny branch lengths and were used in the expression r = K/2T where r = evolutionary rate of the 18S rDNA, K = corrected distances, and T = divergence time (Li and Graur 1991
, p. 67). Our divergence time estimates were calculated using three alternative 18S rDNA evolutionary rates, 0.85%, 2%, and 4% sequence divergence per 100 Myr. The correspondence between Riboclades and previously characterized T. cruzi subgroups, based on D7 rDNA band patterns, mini-exon sequence types, and isoenzyme patterns (zymodemes) is shown in figure 3
. Divergence time-estimate deviations were calculated using the maximum-likelihood error for each branch as estimated using PAML 3.0 (Yang 1997
) (fig. 3
). The basal node in the phylogeny (hypothetical ancestor HA, fig. 3
), separating Riboclade 1 from other subgroups, dates 84.4 MYA, and is in agreement with our previous study (Briones et al. 1999
). Divergence of Riboclade 2 from Riboclades 3 and 4 (hypothetical ancestor HA
, fig. 3
) could be as far as 29.1 Myr, and divergence between Riboclades 3 and 4, 18.9 Myr (hypothetical ancestor HA
, fig. 3 ). Even if a faster rate of evolution is used (2% sequence divergence per 100 Myr) (Ochman and Wilson 1987
; Wilson, Ochman, and Prager 1987
), the divergence would date to 35.9 MYA. This estimate still accommodates the notion that placental mammals that came from island hopping before the faunal exchange in the Pleistocene (2–5 MYA) brought Riboclade 1 to South America (Briones et al. 1999
). At a very fast rate of 4% sequence divergence per 100 Myr, the separation of Riboclade 1 from other Riboclades would be 17.9 Myr. However, the estimation of 18S rDNA evolutionary rate as 0.85% sequence divergence per 100 Myr seems to reflect the rate for the entire gene, whereas the other two alternative rates, as suggested by Escalante and Ayala (1995)
, are biased to highly divergent segments within the gene. If the divergence rate used is 4% sequence divergence per 100 Myr (Moran et al. 1993
), it will agree with recent estimates (18–16 Myr) made using the dihydrofolase reductase and trypanothione reductase genes (Machado and Ayala 2001)
.
|
, and HAConfidence intervals for our divergence time estimates were also calculated using a quartet analysis (Bromham et al. 1998
; Rambaut and Bromham 1998
). Three quartets were chosen: Q1 = (Y, CLBR), (Dm28c, YuYu), Q2 = (MT4167, Sc43cl1), (Colombiana, Silvio X10 cl1), and Q3 = (Y, Peru), (Silvio X10 cl1, YuYu). Quartet pairs' divergence estimates were then calculated using the observed sequence divergence with the fossil calibration (Escalante and Ayala 1995
). The model of sequence evolution used instantaneous substitution probabilities of A–C = 1, A–G = 0.67, A–T = 1, C–G = 1, C–T = 5.51, and G–T = 1, equal base frequencies, and gamma distribution rate parameter = 26.72, as determined by Modeltest as described earlier. The estimates from this model will provide the estimated number of substitutions according to the model and will be corrected by the phylogeny. The divergence of Riboclade 1 from Riboclade 2, first quartet (table 1
), has low and high confidence intervals of 26 and 153 Myr, respectively, and that from quartet 2 has low and high confidence intervals of 21 and 160 Myr, respectively. The divergence of Riboclade 2 from Riboclades 3 and 4, as estimated from quartet 3, has low and high confidence intervals of 14 and 78 Myr, respectively. The three quartets analyzed were not rejected by the chi-square approximation at = 0.05, and our estimates in figure 3
, for these same clades, fall within the range of these confidence intervals. It should be noted that the divergence times entered in the infile are approximations from fossil calibrations and that QDate excludes gapped sites for each quartet. This is a limitation in our analysis, because insertions and deletions are very frequent in the evolution of rDNA sequences, particularly within variable regions corresponding to loops in the secondary structure. Therefore, our estimated divergence dates, actually could be underestimations. Nevertheless, these dates are informative in the sense that the divergence of the T. cruzi groups could be as old as our estimates, and not in the sense of definitive and exact numbers. |
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Phylogenetic analysis of 20 representative strains of T. cruzi based on direct comparison of 18S and D7-24S
rDNA (fig. 1
) confirms previous studies, using other molecular markers, which suggest that T. cruzi can be divided into two main groups or lineages (Souto et al. 1996
; Stothard et al. 1998
; Fernandes et al. 1998a). The Riboclades, as identified in phylogenies in figure 1
, correspond to subgroups characterized by other molecular typing methods (figs. 2 and 3
) (Fernandes et al. 1998b;
Stothard et al. 1998
; Brisse, Barnabe, and Tibayrenc 2000a
). However, these other methods rely on band patterns and other indirect comparisons and not on direct comparison of sequences. Bootstrap support of 98% for the monophyly of Riboclade 1 in the 18S rDNA tree and 100% in the rapidly evolving region D7 in the 24S rDNA tree suggests that the divergence of Riboclade 1 from the other three Riboclades is the primary event (fig. 1
). Considering the low evolutionary rate of the rDNA, it is understandable why this marker can reveal this pattern. Most other markers used for typing of T. cruzi, e.g., multilocus enzyme electrophoresis, randomly amplified polymorphic DNA (RAPD), and restriction fragment length polymorphism, are faster evolving molecules and therefore are more susceptible to saturation and possibly, paralogy, as in the case of RAPD (Melo et al. 1998
). In these cases, the groups of strains might not be identified, or confounded, because the markers evolve too fast to resolve such deep divergences. It is important to notice that Riboclade 1 strains (corresponding to T. cruzi II) were predominantly found in the domestic cycle of T. cruzi transmission, isolated from the circulation of chagasic patients (Zingales et al. 1998
). Also, in a study of 22 patients with monoclonal infections and digestive and cardiac symptoms, it was determined that infective strains belong to T. cruzi II (E. Lages-Silva, E. Chiari, A. M. Macedo, personal communication). Strains of T. cruzi I (Riboclade 2) and zymodeme 3 (Riboclade 4) have been associated with the sylvatic cycle of the parasite and have been mainly isolated from opossums, wild rodents, and triatomines (Miles and Cibulskis 1986
; Zingales et al. 1998
; Fernandes et al. 2001
). Isolates of this group were also obtained from chagasic patients in the acute or chronic phases of the disease. The implication of these observations is that although Riboclade 2 can be found in humans when typing is based on parasites isolated from peripheral blood, it might not be the cause of heart and digestive tissue damage and disease. It has been extensively shown that human patients are almost always infected with multiclonal strains, and that within the same patient, genetic heterogeneity is verified when strains isolated from the blood or from the affected organs are compared (Vago et al. 2000)
.
The analysis presented here also suggests that Riboclade 2 is a sister group of the Riboclades that contain representatives of zymodeme Z3 (Miles et al. 1978
) and group 1/2 (Souto et al. 1996
). The subtree (Riboclade 2 [Riboclade 3, Riboclade 4]) (fig. 1A
) has an internal divergence which can be noticed also in the D7-24S rDNA tree (fig. 1B
). This contrasts with the previous proposition that the Riboclade 1 group has more variability than Riboclade 2 (Brisse, Barnabe, and Tibayrenc 2000a
).
Possibly, the variability found among Riboclades 2, 3, and 4 might be explained by the evolution and distribution of specific mammalian hosts, and of the insect vectors, reduviid bugs, that evolved in South America during the Cenozoic (Marcilla et al. 2001)
. Group 1/2, here represented by isolates NRcl3 and S03cl5, is an interesting group of strains. This group has been characterized by Souto et al. (1996)
based on the structure of the rDNA cistrons, the intergenic region of the mini-exon gene, and RAPD profiles. We demonstrated that the genome of these isolates contains two types of rDNA cistrons (1 and 2) and that copies of type 2 are 10-fold more abundant than copies of type 1 (Souto et al. 1996
). We have also observed that only type 2 rRNA genes are transcribed in group 1/2 strains. In the present study, we compared the D7-24S rDNA sequences of the two types of rDNA cistrons of SO3cl5. In the phylogeny shown in figure 1B
, we observe that D7-24S rDNA type 1 fragment (SO3cl5[1]) branched along with other Riboclade 1 sequences, whereas D7-24S rDNA type 2 fragment (SO3cl5 [2]) branched along with Riboclade 3 sequences, which also contains parasites that have only the D7-24S rDNA type 2 gene. Accordingly, the analysis of the 18S rRNA gene of type 2 cistron of NRcl3, also belonging to group 1/2, places this sequence in Riboclade 3 in the phylogeny shown in figure 1A.
The origin of group 1/2 is still a matter of debate. Analysis of the structure of the intergenic region of the mini-exon gene and that of 50–60 anonymous loci generated by RAPD places this group of strains in lineage 1 (corresponding to ribodeme 1) (Souto et al. 1996
). In that study we hypothesized that group 1/2 could have been originated by an eventual transfer of rDNA from individuals of group 2 to group 1, either by mating or by horizontal gene transfer. Analysis of other genetic markers would indicate if other sets of genes (besides rDNA) have also been transferred in group 1/2. Supporting this hypothesis, three reports suggest the formation of hybrid organisms in sylvatic T. cruzi populations (Carrasco et al. 1996
), in sympatric clinical isolates (Bogliolo, Lauria-Pires, and Gibson 1996
), and in experiments, as dual-resistant T. cruzi biological clones could be obtained following copassage of parents carrying single episomal drug-resistant markers (Stothard, Frame, and Miles 1999
). Recently, Machado and Ayala (2001) concluded
, based on nucleotide sequences from two nuclear genes that encode the enzymes trypanothione reductase and dihydrofolate reductase-thymidilate synthetase and a region from the mitochondrial genome, that two widely distributed isoenzyme types—one corresponding to group 1/2—are hybrids.
Typing based on D7-24S rDNA was improved also with the identification of D7-24S rDNA type 3. Because of a small difference in size between the 125- and the 117-bp D7-24S rDNA amplicons, some strains originally typed as rDNA type 1 (Souto et al. 1996
) represent another group corresponding to our Riboclade 4.
In conclusion, in the absence of fossil records of trypanosomatids, the only way to estimate the divergence times is through molecular data. Therefore, we have tested the molecular clock in the 18S rDNA and identified sequence subsets which reconstruct phylogenies that can be used to estimate divergence times because they passed the likelihood ratio test, and an evolutionary rate for the entire 18S rDNA has been estimated (Escalante and Ayala 1995
). Our estimates of divergence times were, therefore, calibrated by other taxonomic groups whose respective divergence times were dated from the fossil record (Escalante and Ayala 1995
). Other methods to date and estimate confidence intervals were proposed, although they require that quartets in the data set are dated by the fossil record (Rambaut and Bromham 1998
). The fossil record estimates should be taken in the perspective that these estimates indicate that a particular group is as old as the estimate but do not exclude that it can be older. Variations of evolutionary rates among lineages are tested by the likelihood ratio test, and the phylogeny presented in figure 3 passed the test. Another problem is that the 18S rDNA could be evolving in the whole T. cruzi clade faster than observed for the other groups from which the rate of 0.85% sequence divergence per 100 Myr was obtained. Because of this problem we estimated our divergence times also using two other rates (2% and 4% sequence divergence per 100 Myr) (Ochman and Wilson 1987
; Wilson, Ochman, and Prager 1987
; Moran et al. 1993
), which accommodate for a 2.4-and 4.7-fold rate increase, respectively. Even with a 2.4-fold rate increase, the divergence would date back to approximately 36 MYA, when rodents and primates first entered into South America (Simpson 1980
). Our proposition is that Riboclades 2, 3, and 4 were evolving in South American marsupials, isolated from North American placentals, parasitized by Riboclade 1, during most of the Cenozoic period, which corresponds approximately to the divergence time estimated here (fig. 4
) (Penny and Hasegawa 1997
). The hypothetical ancestor of Riboclades 1, 2, 3, and 4 (fig. 3
) was spread over South and North America 84.4 MYA, still during the period when the geographical isolation of South America from the Old World was not a barrier to mammalian migration. After the isolation of South America in the Cenozoic, its mammalian fauna was composed of marsupials, edentates, and ungulates. Interestingly, T. cruzi does not infect large ungulate mammals, such as cows and horses (Wendel et al. 1992
), which possibly restricted T. cruzi in South America to marsupial and edentate hosts. According to this hypothesis, T. cruzi Riboclade 1 might have entered South America either during the Oligocene, along with rodents and primates that came via island hopping, or during the great faunal exchange during the Pliocene-Pleistocene, when South and North America were connected via the Panama isthmus. In North America the association between Riboclade 1 and placental mammals and Riboclade 2 and marsupials is very strong and might be evidence for independent evolution of these two clades for a long period (Clark and Pung 1994
). The North American marsupial Didelphis virginiana is of South American origin; possibly ancestors of these mammals introduced T. cruzi Riboclade 2 into North America during the faunal exchange (Briones et al. 1999
). In our study we are trying to estimate events that are related to the speciation time, which is not exact, and the very slow separation of South America from Africa. This is estimated to have occurred during the late Jurassic into the late Cretaceous, and even in the late Cretaceous (approximately 100 Myr) there was a narrow seaway between Africa and South America (Futuyma 1998
, p. 182). An alternative is to assume a faster rate of evolution (4% divergence per 100 Myr) for the 18S rDNA in T. cruzi. This would suggest that the divergence of strains occurred recently (less than 18 Myr) and that it might be a result of either coevolution with insect vectors or geographical barriers within South America that isolated marsupials and placental mammals carrying T. cruzi II. This latter scenario is favored by our estimates (table 2 in Briones et al. 1999
) and recent estimates using dihydrofolate reductase and trypanothione reductase, with a superimposed divergence between T. cruzi and T. brucei of 100 Myr (Machado and Ayala 2001)
. This more recent study argues that if the divergence is as deep as 88–37 MYA, the silent substitution rates for these two protein-coding genes would be an order of magnitude smaller than the average synonymous substitution rates of other organisms (Machado and Ayala 2001)
. However, codon bias and expression levels of particular protein-coding genes are known to affect the rate of synonymous substitutions in such orders of magnitude (Akashi 1
994). In fact, the very broad range of confidence intervals of our estimates, as calculated by quartet analysis (QDate) (table 2
), still do not allow us to decide if the divergence among T. cruzi strains occurred 18–16 MYA (Briones et al. 1999
; Machado and Ayala 2001
) or 88–37 MYA (Briones et al. 1999
).
|
|
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
We thank Paulo B. Paiva and Dr. Adriana Brunstein for manuscript review and Beatriz Schnabel for technical assistance. S.Y.K. and G.F.O.S. received, respectively, undergraduate and graduate fellowships from CNPq (Brazil). B.Z. received grants from FAPESP (Brazil) and M.R.S.B received grants from FAPESP (Brazil) and an International Research Scholar grant from the Howard Hughes Medical Institute (USA).
|
Footnotes |
|---|
Kenneth Wolfer, Reviewing Editor
Keywords: Chagas' disease
Trypanosoma cruzi
evolution
rRNA gene
molecular clock
maximum-likelihood 
Address for correspondence and reprints: Marcelo R. S. Briones, Disciplina de Microbiologia, Escola Paulista de Medicina, Universidade Federal de São Paulo, Rua Botucatu, 862, 3° andar, CEP 04023-062, São Paulo, S.P., Brazil. marcelo{at}ecb.epm.br
.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Akashi H., 1994 Synonymous codon usage in Drosophila melanogaster: natural selection and translation accuracy Genetics 136:927-935[Abstract]
Bogliolo A. R., L. Lauria-Pires, W. C. Gibson, 1996 Polymorphisms in Trypanosoma cruzi: evidence of genetic recombination Acta Trop 61:31-40[Web of Science][Medline]
Briones M. R. S., R. P. Souto, B. Stolf, B. Zingales, 1999 The evolution of two Trypanosoma cruzi subgroups inferred from rRNA genes can be correlated with the interchange of American mammalian faunas in the Cenozoic and has implications to pathogenicity and host specificity Mol. Biochem. Parasitol 104:219-232[Web of Science][Medline]
Brisse S., C. Barnabe, M. Tibayrenc, 2000a. Identification of six Trypanosoma cruzi phylogenetic lineages by random amplified polymorphic DNA and multilocus enzyme electrophoresis Int. J. Parasitol 30:35-44[Web of Science][Medline]
Brisse S., J. C. Dujardin, M. Tibayrenc, 2000b. Identification of six Trypanosoma cruzi lineages by sequence-characterised amplified region markers Mol. Biochem. Parasitol 111:95-105[Web of Science][Medline]
Bromham L., A. Rambaut, R. Fortey, A. Cooper, D. Penny, 1998 Testing the Cambrian explosion hypothesis by using a molecular dating technique Proc. Natl. Acad. Sci. USA 95:12386-12389
Carrasco H. J., I. A. Frame, S. A. Valente, M. A. Miles, 1996 Genetic exchange as a possible source of genomic diversity in sylvatic populations of Trypanosoma cruzi Am. J. Trop. Med. Hyg 54:418-424
Clark C. G., O. J. Pung, 1994 Host specificity of ribosomal DNA variants in sylvatic Trypanosoma cruzi from North America Mol. Biochem. Parasitol 66:175-179[Web of Science][Medline]
Escalante A. A., F. J. Ayala, 1995 Evolutionary origin of Plasmodium and other apicomplexa based on rRNA genes Proc. Natl. Acad. Sci. USA 92:5793-5797
Felsenstein J., 1981 Evolutionary trees from DNA sequences: a maximum-likelihood approach J. Mol. Evol 17:368-376[Web of Science][Medline]
———. 1988 Phylogenies from molecular sequences: inference and reliability Annu. Rev. Genet 22:521-565[Web of Science][Medline]
Fernandes O., R. H. Mangia, C. A. Lisboa, A. P. Pinho, C. M. Morel, B. Zingales, D. A. Campbell, A. M. Jansen, 1999 The complexity of the sylvatic cycle of Trypanosoma cruzi in Rio de Janeiro state (Brazil) revealed by the non-transcribed spacer of the mini-exon gene Parasitology 118:161-166
Fernandes A. P., K. Nelson, S. M. Beverley, 1993 Evolution of nuclear ribosomal RNAs in kinetoplastid protozoa: perspectives on the age and origins of parasitism Proc. Natl. Acad. Sci. USA 90:11608-11612
Fernandes O., S. S. Santos, E. Cupolillo, et al. (12 co-authors) 2001 Trypanosomiasis in the Brazilian Amazon—a mini-exon multiplex PCR to distinguish the major groups of Trypanosoma cruzi and Trypanosoma rangeli Trans. R. Soc. Trop. Med. Hyg 95:1-3[Web of Science][Medline]
Fernandes O., R. P. Souto, J. A. Castro, et al. (12 co-authors) 1998a. Brazilian isolates of Trypanosoma cruzi from humans and triatomines classified into two lineages using mini-exon and ribosomal RNA sequences Am. J. Trop. Med. Hyg 58:807-811[Abstract]
Fernandes O., N. R. Sturm, R. Derr, D. A. Campbell, 1998b. The mini-exon gene: a genetic marker for zymodeme III of Trypanosoma cruzi Mol. Biochem. Parasitol 95:129-133[Web of Science][Medline]
Futuyma D. J., 1998 Evolutionary Biology Sinauer Associates, Inc., Sunderland, Mass.
Galtier N., M. Gouy, C. Gautier, 1996 SEAVIEW and PHYLO_WIN: two graphic tools for sequence alignment and molecular phylogeny Comput. Appl. Biosci 12:543-548
Gomes M. L., A. M. Macedo, S. D. Pena, E. Chiari, 1998 Genetic relationships between Trypanosoma cruzi strains isolated from chronic chagasic patients in southern Brazil as revealed by RAPD and SSR-PCR analysis Acta Trop 69:99-109[Web of Science][Medline]
Gordon D., C. Abajian, P. Green, 1998 Consed: a graphical tool for sequence finishing Genome Res 8:195-202
Hasegawa M., H. Kishino, T. A. Yano, 1985 Dating of the human-ape splitting by a molecular clock of mitochondrial DNA J. Mol. Evol 22:160-174[Web of Science][Medline]
Hillis D. M., B. K. Mable, C. Moritz, 1996 Applications of molecular systematics: the state of the field and a look to the future Pp. 515–543. in D. M. Hillis, C. Moritz, and B. K. Mable, eds. Molecular systematics. Sinauer Associates, Inc., Sunderland, Mass
Jukes T. H., C. R. Cantor, 1969 Evolution of protein molecules Pp. 21–123 in H. N. Munro, ed. Mammalian protein metabolism. Academic Press, New York
Kimura M., 1980 A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleotide sequences J. Mol. Evol 16:111-120[Web of Science][Medline]
———. 1981 Estimation of evolutionary distances between homologous nucleotide sequences Proc. Natl. Acad. Sci. USA 78:454-458
Li W. H., D. Graur, 1991 Fundamentals of molecular evolution Sinauer Associates, Sunderland, Mass
Machado C. A., F. J. Ayala, 2001 Nucleotide sequences provide evidence of genetic exchange among distantly related lineages of Trypanosoma cruzi Proc. Natl. Acad. Sci. USA 96:7396-7401
Marcilla A., M. D. Bargues, J. M. Ramsey, E. Magallon-Gastelum, P. M. Salazar-Schettino, F. Abad-Franch, J. Dujardin, C. J. Schofield, S. Mas-Coma, 2001 The ITS-2 of the nuclear rDNA as a molecular marker for populations, species, and phylogenetic relationships in Triatominae (Hemiptera: Reduviidae), vectors of Chagas disease Mol. Phylogenet. Evol 18:136-142[Web of Science][Medline]
Melo A. S. A., L. P. Almeida, A. L. Colombo, M. R. S. Briones, 1998 Evolutionary distances and identification of Candida species in clinical isolates by randomly amplified polymorphic DNA (RAPD) Mycopathologia 142:57-66[Web of Science][Medline]
Mendonça M. B. A., N. S. Nehme, S. S. Santos, et al. (11 co-authors) 2001 Two main clusters within Trypanosoma cruzi zymodeme3 are defined by distinct regions of the ribosomal RNA cistron Parasitology (in press)
Miles M., R. E. Cibulskis, 1986 Zymodeme characterization of Trypanosoma cruzi Parasitol. Today 4:94-97
Miles M. A., A. Souza, M. Povoa, J. J. Shaw, R. Lainson, P. E. Toy, 1978 Isozymic heterogeneity of Trypanosoma cruzi in the first authochthonous patients with Chagas' disease in Amazonian Brazil Nature 272:819-821[Medline]
Moran N. A., M. A. Munson, P. Baumann, H. Ishikawa, 1993 A molecular clock in endosymbiotic bacteria is calibrated using the insect hosts Proc. R. Soc. Lond. Ser. B 254:167-171
Morel C., E. Chiari, E. P. Camargo, D. M. Mattei, A. J. Romanha, L. Simpson, 1980 Strains and clones of Trypanosoma cruzi can be characterized by pattern of restriction endonuclease products of kinetoplast DNA minicircles Proc. Natl. Acad. Sci. USA 77:6810-6814
Ochman H., A. C. Wilson, 1987 Evolution in bacteria: evidence for a universal substitution rate in cellular genomes J. Mol. Evol 26:74-86[Web of Science][Medline]
Penny D., M. Hasegawa, 1997 The platypus put in its place Nature 387:549-550[Medline]
Posada D., K. A. Crandall, 1998 Modeltest: testing the model of DNA substitution Bioinformatics 14:817-818
Rambaut A., L. Bromham, 1998 Estimating divergence dates from molecular sequences Mol. Biol. Evol 15:442-448[Abstract]
Satellite-Meeting. 1999 Recommendations from an International Symposium to commemorate the 90th anniversary of the discovery of Chagas disease. April 11–16 1999, Rio de Janeiro, Brazil Mem. Inst. Oswaldo Cruz 94:429-432
Simpson G. G., 1980 Splendid isolation The curious history of South American mammals. Yale University Press, New Haven, Conn
Souto R. P., O. Fernandes, A. M. Macedo, D. A. Campbell, B. Zingales, 1996 DNA markers define two major phylogenetic lineages of Trypanosoma cruzi Mol. Biochem. Parasitol 83:141-152[Web of Science][Medline]
Souto R. P., B. Zingales, 1993 Sensitive detection and strain classification of Trypanosoma cruzi by amplification of a ribosomal RNA sequence Mol. Biochem. Parasitol 62:45-52[Web of Science][Medline]
Stevens J., H. A. Noyes, W. C. Gibson, 1998 The evolution of trypanosomes infecting humans and primates Mem. Inst. Oswaldo Cruz 93:669-676[Web of Science][Medline]
Stothard J. R., I. A. Frame, M. A. Miles, 1999 Genetic diversity and genetic exchange in Trypanosoma cruzi: dual drug-resistant "progeny" from episomal transformants Mem. Inst. Oswaldo Cruz 94: (Suppl. I) 189-193
Stothard J. R., I. A. Frame, H. J. Carrasco, M. A. Miles, 1998 On the molecular taxonomy of Trypanosoma cruzi using riboprinting Parasitology 117:243-247
Swofford D. L., 1998 PAUP*: phylogenetic analysis using parsimony (* and other methods). Version 4b6 Sinauer Associates, Sunderland, Mass
Swofford D. L., G. J. Olsen, P. J. Waddell, D. M. Hillis, 1996 Phylogenetic inference Pp. 407–514 in D. M. Hillis, C. Moritz, and B. K. Mable, eds. Molecular Systematics. Sinauer Associates, Sunderland, Mass
Tamura K., M. Nei, 1993 Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees Mol. Biol. Evol 10:512-526[Abstract]
Vago A. R., L. O. Andrade, A. A. Leite, D. d'Avila Reis, A. M. Macedo, S. J. Adad, S. Tostes, M. C. Moreira, G. B. Filho, S. D. Pena, 2000 Genetic characterization of Trypanosoma cruzi directly from tissues of patients with chronic Chagas disease: differential distribution of genetic types into diverse organs Am. J. Pathol 156:1805-1809
Wendel S., Z. Brener, M. E. Camargo, A. Rassi, 1992 Chagas Disease (American trypanosomiasis): its impact on transfusion and clinical medicine ISBT Brazil'92—SBHH (Sociedade Brasileira de Hematologia e Hemoterapia), São Paulo, Brazil
Wilson A. C., H. Ochman, E. M. Prager, 1987 Perspectives molecular time scale for evolution Trends Genet 3:241-247
Yang Z., 1996 Maximum-likelihood models for combined analyses of multiple sequence data J. Mol. Evol 42:587-596[Web of Science][Medline]
———. 1997 PAML: a program package for phylogenetic analysis by maximum-likelihood CABIOS 13:555-556
Zingales B., R. P. Souto, R. H. Mangia, C. V. Lisboa, D. A. Campbell, J. R. Coura, A. Jansen, O. Fernandes, 1998 Molecular epidemiology of American trypanosomiasis in Brazil based on dimorphisms of rRNA and mini-exon gene sequences Int. J. Parasitol 28:105-112[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. M. Pennington, C. Paiz, L. M. Grajeda, and C. Cordon-Rosales Concurrent Detection of Trypanosoma cruzi Lineages I and II in Domestic Triatoma dimidiata from Guatemala Am J Trop Med Hyg, February 1, 2009; 80(2): 239 - 241. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. P. A. Martins, A. Marcili, R. E. P. Castanho, A. L. S. Therezo, J. C. P. de Oliveira, R. B. Suzuki, M. M. G. Teixeira, J. A. da Rosa, and M. A. Speranca Rural Triatoma rubrovaria from Southern Brazil Harbors Trypanosoma cruzi of Lineage IIc Am J Trop Med Hyg, September 1, 2008; 79(3): 427 - 434. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J. Westenberger, C. Barnabe, D. A. Campbell, and N. R. Sturm Two Hybridization Events Define the Population Structure of Trypanosoma cruzi Genetics, October 1, 2005; 171(2): 527 - 543. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. E. Steiper, N. M. Young, and T. Y. Sukarna Genomic data support the hominoid slowdown and an Early Oligocene estimate for the hominoid-cercopithecoid divergence PNAS, December 7, 2004; 101(49): 17021 - 17026. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 PATTERNS OF CALIBRATION AGE SENSITIVITY WITH QUARTET DATING METHODS Journal of Paleontology, January 1, 2004; 78(1): 7 - 30.
|
|
|
|
 |
|
 J. M. Di Noia, C. A. Buscaglia, C. R. De Marchi, I. C. Almeida, and A. C.C. Frasch A Trypanosoma cruzi Small Surface Molecule Provides the First Immunological Evidence that Chagas' Disease Is Due to a Single Parasite Lineage J. Exp. Med., February 11, 2002; 195(4): 401 - 413. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||