Defining orthologous groups among multicopy genes prior to inferring phylogeny, with special emphasis on the Triticeae (Poaceae)

Diversity within the genus Elymus (Poaceae: Triticeae) II: analyses of variation within 5S nrDNA restrict membership in the genus to species with StH genomes

The genus Elymus is a repository for a large number of species that have been difficult to classify by traditional techniques due to their remarkable levels of polymorphism. Following the genome analyses of Yen and Yang (Genus Elymus 5:58-362, 2013), we used sequences of the nr5SDNA to investigate diversity within those 24 species having St and H haplomes (Baum et al. Mol Genet Genomics 290:329-42, 2015) and for which the genome status was known. The present work extends this analysis to include eight species for which there was no information on genomic status. Our results show that these eight have nr5SDNA sequences that can be assigned to unit classes of orthologous sequences found in St and H haplomes, suggesting that the presence of St and H haplomes is characteristic of the genus. We then carried out a set of canonical discriminant analyses based on 247 DNA new sequences from these 8 species plus the 1054 sequences previously identified from 24 Elymus species. Sequences were analyzed to answer the following questions: Do the species integrate or are they different? Are the tetraploids different from the higher-ploid species? Are the species united within sections, or the same within regions? How do the species fare when divided according to sections? The main results of the canonical discriminant analyses are that the species are united within the tetraploids and within the hexaploids, within each region and within each section. In addition, a series of classificatory discriminant analyses showed that the identification tests are different, although not sufficiently useful for the discrimination of all the species. We also demonstrate the power of our approach by showing that the voucher for Elymus mobilis is not Elymus at all, but Leymus.

Diversity within the genus Elymus (Poaceae: Triticeae) as investigated by the analysis of the nr5S rDNA variation in species with St and H haplomes

The genus Elymus ("Ryegrass") is a repository for a range of species with a variety of haplome contents; hence the pejorative name "dustbin" genus. We have analyzed 1,059 sequences from 128 accessions representing 24 species to investigate the relationships among the StH haplomes-containing species described by Yen and Yang (Genus Elymus Beijing 5:58-362, 2013). Sequences were assigned to "unit classes" of orthologous sequences and subjected to a suite of analyses including BLAST (Basic Local Alignment Search Tool) searches, phylogenetic analysis and population genetic analysis to estimate species diversity. Our results support the genome analyses in Yen and Yang (Genus Elymus Beijing 5:58-362, 2013), i.e., genomic constitution StStHH including variants restricted to Elymus. Population genetic analysis of the 5S nrDNA sequence data revealed that the within-species variance component is roughly ±89 %; thus, we were unable to identify molecular markers capable to separate the 24 species analyzed. Separate phylogenetic analyses of the two unit classes and of all the data exhibit a trend only of the species to cluster on the phylograms. Finally, the analysis provides evidence for the multiple origins of American and Eurasian species.

Polyploidy and hybridization as main factors of speciation: complex reticulate evolution within the grass genus Helictochloa

To study the origin and evolution of naturally occurring polyploids, we performed phylogenetic analyses of nuclear ribosomal DNA spacers combined with molecular cytogenetics in 55 accessions of 27 taxa of the oat genus Helictochloa. A complex pattern of reticulate evolution was revealed with many diploid species and extensive polyploidy up to 20x. Altogether 11 groups of internal transcribed spacer (ITS) sequences can be distinguished. Sequences from 1-3 different ITS lineages were detected in polyploids. Cytogenetic data allow reconstruction of 8 basic monoploid chromosome sets. Six of these genomes occur in different combinations in the polyploid species. Two genomes are only found in diploids. Our sequence and karyological data highlight the occurrence of autopolyploidy and allopolyploidy, provide new information about the evolutionary history of taxa, and allow a more accurate systematic treatment of the concerned species. The geographical distribution of the 11 ITS lineages distinguished is highly structured and points to an origin of the genus in western Asia, presumably in grasslands like steppes or mountain steppes and meadows. The evolutionary basal lineages are of Asian, Minor Asian and east Mediterranean distribution and are present also in North America. The western and central parts of the Mediterranean and northern Europe harbor the modern lineages.

What does the nr5S DNA multigene family tell us about the genomic relationship between Dasypyrum breviaristatum and D. villosum (Triticeae: Poaceae)?

The genus Dasypyrum contains two species: the annual and widespread D. villosum (2x = 2n = 14) and the perennial and generally rare D. breviaristatum (2x = 2n = 14 and 4x = 2n = 28). The origin of the latter and its genome constitution have been subject of several studies. There is agreement that the genome of the diploid D. villosum (VV) is different from the diploid cytotype of D. breviaristatum (VbVb), but there is no agreement of the constitution of the tetraploid cytotype, specifically whether is it an autotetraploid or an allotetraploid. This is a long-standing disagreement that this study aims to resolve using the 5S nrDNA as a genomic marker. Our studies suggest that the 4x D. breviaristatum is an allotetraploid (VVVbVb).

What does the 5S rRNA multigene family tell us about the origin of the annual Triticeae (Poaceae)?

We have investigated the complex relationships among the annual genera within the tribe Triticeae through phylogenetic analyses of the 5S rRNA multigene family. Cloned sequences were assigned to groups of orthologous sequences, called unit classes, that were subjected to several analyses including BLAST (Basic Local Alignment Search Tool) searches to assess possible ancestral relationships with perennial genera; phylogenetic analyses using parsimony (Pars), maximum likelihood (ML), and Bayesian methods; and minimum reticulation networks from the Pars, ML, and Bayesian trees. In this study, we included genera with both annual and perennial species, such as Dasypyrum, Hordeum, and Secale. BLAST pointed to Pseudoroegneria (carrier of the St genome) and possibly Thinopyrum (carrier of the J genome) as the potential next of kin. However, Thinopyrum and Pseudoroegneria have never fallen together on the individual trees with the former generally associated with Crithopsis, Aegilops, Triticum, and Dasypyrum, while the latter is usually associated with the rest of the genera within Triticeae. The "long" unit classes placed Dasypyrum breviaristatum together with Dasypyrum villosum, whereas the "short" unit classes put them far apart on the trees. None of the gene trees alone was able to summarize the complex relationships among the genera, in line with previous results in the Triticeae. However, the application of tools designed to display phylogenetic networks was able to depict the complex links among the genera based on the short and the long gene trees, including the close link between Thinopyrum and Pseudoroegneria suggested by the phylogenetic analyses. In addition, our analyses provide support for the hypothesis that at least some annual Triticeae taxa are derived from their perennial relatives.

Phylogenetic relationships among the polyploid and diploid Aegilops species inferred from the nuclear 5S rDNA sequences (Poaceae: Triticeae)

Phylogenetic inferences of the polyploid Aegilops taxa were drawn based upon the analysis of 909 nuclear 5S rDNA sequences obtained from 15 Aegilops polyploid taxa (531 sequences new to this paper) and 378 sequences from our previous study on the diploid taxa. The 531 sequences can be split into two orthologous groups (unit classes), the long AE1 and short AE1 previously identified in the diploid set. An examination of the relationships between unit classes and their associated haplomes suggests that U haplome sequences found in Ae. umbellulata are the closest to the T sequences found in Amblyopyrum muticum and that sequences of the polyploid species expected to be the M type found in Ae. comos are more similar to the T haplome sequences, except in the three hexaploids Ae. glumiaristata, Ae. juvenalis, and Ae. vavilovii and the tetraploid Ae. crassa where they are found to be similar to the M haplome sequences. These three hexaploid taxa likely originated from the tetraploid Ae. crassa (DM), while the closest taxon to the fourth hexaploid, Ae. recta, is the tetraploid Ae. neglecta (UM). Based upon the distribution of the unit classes, several reticulate phylogenies depicting evolutionary relationships among diploid, tetraploid, and hexaploid taxa were constructed; however, none of these widely used methods could depict the expected reticulate relationship as previously drawn from cytogenetic analyses in this group of allopolyploid species. These results suggest that evolutionary relationships derived from models based upon the assumption of bifurcating species require careful interpretation when these same models are applied to species with reticulate evolution.

Plant DNA sequencing for phylogenetic analyses: from plants to sequences

DNA sequences are important sources of data for phylogenetic analysis. Nowadays, DNA sequencing is a routine technique in molecular biology laboratories. However, there are specific questions associated with project design and sequencing of plant samples for phylogenetic analysis, which may not be familiar to researchers starting in the field. This chapter gives an overview of methods and protocols involved in the sequencing of plant samples, including general recommendations on the selection of species/taxa and DNA regions to be sequenced, and field collection of plant samples. Protocols of plant sample preparation, DNA extraction, PCR and cloning, which are critical to the success of molecular phylogenetic projects, are described in detail. Common problems of sequencing (using the Sanger method) are also addressed. Possible applications of second-generation sequencing techniques in plant phylogenetics are briefly discussed. Finally, orientation on the preparation of sequence data for phylogenetic analyses and submission to public databases is also given.

Elimination of 5S DNA unit classes in newly formed allopolyploids of the genera Aegilops and Triticum

Two classes of 5S DNA units, namely the short (containing units of 410 bp) and the long (containing units of 500 bp), are recognized in species of the wheat (the genera Aegilops and Triticum) group. While every diploid species of this group contains 2 unit classes, the short and the long, every allopolyploid species contains a smaller number of unit classes than the sum of the unit classes of its parental species. The aim of this study was to determine whether the reduction in these unit classes is due to the process of allopolyploidization, that is, interspecific or intergeneric hybridization followed by chromosome doubling, and whether it occurs during or soon after the formation of the allopolyploids. To study this, the number and types of unit classes were determined in several newly formed allotetraploids, allohexaploids, and an allooctoploid of Aegilops and Triticum. It was found that elimination of unit classes of 5S DNA occurred soon (in the first 3 generations) after the formation of the allopolyploids. This elimination was reproducible, that is, the same unit classes were eliminated in natural and synthetic allopolyploids having the same genomic combinations. No further elimination occurred in the unit classes of the 5S DNA during the life of the allopolyploid. The genetic and evolutionary significance of this elimination as well as the difference in response to allopolyploidization of 5S DNA and rDNA are discussed.

Codependence of repetitive sequence classes in genomes: phylogenetic analysis of 5S rDNA families in Hordeum (Triticeae: Poaceae)

To complete our study of the genus Hordeum and to elaborate a phylogeny of species based upon 5S rDNA sequences, we have cloned and sequenced PCR amplicons from seven American polyploid species to generate 164 new 5S rRNA gene sequences. These sequences were analysed along with the more than 2000 5S rDNA sequences previously generated from the majority of species in Hordeum to provide a comprehensive picture of the distribution (presence or absence) of 5S rDNA unit classes (orthologous groups) in this genus as well as insights into the phylogeny of Hordeum. Testing of substitution models for each unit class based upon the consensus sequences of all the taxa as well as for each unit class within the genus found that the general best fit was TPM3uf+G, from which a maximum-likelihood tree was calculated. A novel application of cophylogenetic analysis, where relationships among unit classes were treated as host-parasite interactions, depicted some significant pair links under tests of randomness indicative of nonrandom codivergence among several unit classes within the same taxon. The previous classification of four genomic groups is reflected in combinations of unit classes, and it is proposed that current taxa developed from ancient diploidized paleopolyploids and that some were subjected to gene loss, i.e., unit class loss. Finally, separate phylogenetic analyses performed for the tetraploid and hexaploid species were used to derive a working model describing the phylogeny of the polyploid taxa from their putative diploid ancestry.

Multiple patterns of rDNA evolution following polyploidy in Oryza

Ribosomal ITS sequences are commonly used for phylogenetic reconstruction because they are included in rDNA repeats, and hence are ubiquitous and present in high copy number. Ribosomal rDNA repeats often undergo rapid concerted evolution within and between arrays. Interspecific hybridization merges divergent repeat types in a single nucleus, setting in motion evolutionary processes leading to coexistence, maintenance of paralogs, origin of novel sequence variants, loss of arrays, or inter-array sequence homogenization via concerted evolution. Here we examined ITS polymorphism within and among six Oryza tetraploids of varying genomic composition to infer the extent and direction of concerted evolution following allopolyploid speciation. We demonstrate that different polyploids have experienced varying fates, including maintenance or homogenization of divergent arrays, even among allopolyploids having the same genomic origins but in different geographic locations. Bidirectional concerted evolution, in which arrays become homogenized to alternative progenitor diploid types in different allopolyploid derivatives, is evident among species in one clade. Our results exemplify the panoply of outcomes for ribosomal DNA evolution following allopolyploid speciation.

Loss of 5S rDNA units in the evolution of Agropyron, Pseudoroegneria, and Douglasdeweya

We have investigated relationships among the three closely related genera Agropyron, Pseudoroegneria, and Douglasdeweya. Based upon grouping of 330 5S rDNA sequences into unit classes, we found that Douglasdeweya, with the genomic constitution PPStSt, has 2 unit classes, the long P1 and short S1, and Pseudoroegneria, with the genomic constitution StSt or StStStSt, has the long S1 and short S1 unit classes. In contrast, only the long P1 unit class was found in species of the genus Agropyron (PP). Having a single unit class is unique among all the genera of the tribe Triticeae investigated so far and may reflect gene loss or lineage sorting during its genesis. The presence of the short S1 and long P1 unit classes confirms the amphiploid origin of Douglasdeweya.

Molecular diversity of the 5S rRNA gene and genomic relationships in the genus Avena (Poaceae: Aveneae)

The molecular diversity of the rDNA sequences (5S rDNA units) in 71 accessions from 26 taxa of Avena was evaluated. The analyses, based on 553 sequenced clones, indicated that there were 6 unit classes, named according to the haplomes (genomes) they putatively represent, namely the long A1, long B1, long M1, short C1, short D1, and short M1 unit classes. The long and short M1 unit classes were found in the tetraploid A. macrostachya, the only perennial species. The long M1 unit class was closely related to the short C1 unit class, while the short M1 unit class was closely related to the long A1 and long B1 unit classes. However, the short D1 unit class was more divergent from the other unit classes. There was only one unit class per haplome in Avena, whereas haplomes in the Triticeae often have two. Most of the sequences captured belonged to the long A1 unit class. Sequences identified as the long B1 unit class were found in the tetraploids A. abyssinica and A. vaviloviana and the diploids A. atlantica and A. longiglumis. The short C1 unit class was found in the diploid species carrying the C genome, i.e., A. clauda, A. eriantha, and A. ventricosa, and also in the diploid A. longiglumis, the tetraploids A. insularis and A. maroccana, and all the hexaploid species. The short D1 unit class was found in all the hexaploid species and two clones of A. clauda. It is noteworthy that in previous studies the B genome was found only in tetraploid species and the D genome only in hexaploid species. Unexpectedly, we found that various diploid Avena species contained the B1 and D1 units. The long B1 unit class was found in 3 accessions of the diploid A. atlantica (CN25849, CN25864, and CN25887) collected in Morocco and in 2 accessions of A. longiglumis (CIav9087 and CIav9089) collected in Algeria and Libya, respectively, whereas only 1 clone of A. clauda (CN21378) had the short D1 unit. Thus there might be a clue as to where to search for diploids carrying the B and D genomes. Avena longiglumis was found to be the most diverse species, possibly harboring the A, B, and C haplomes. The long M1 and short M1 are the unit classes typical of A. macrostachya. These results could explain the roles of A. clauda, A. longiglumis, and A. atlantica in the evolution of the genus Avena. Furthermore, one clone of the tetraploid A. murphyi was found to have sequences belonging to the short D1 unit class, which could indicate that A. murphyi might have been the progenitor of hexaploid oats and not, as postulated earlier, A. insularis. The evolution of Avena did not follow the molecular clock. The path inferred is that the C genome is more ancient than the A and B genomes and closer to the genome of A. macrostachya, the only existing perennial, which is presumed to be the most ancestral species in the genus.

Molecular evolution of satellite DNA repeats and speciation of lizards of the genus Darevskia (Sauria: Lacertidae)

Satellite DNA repeats were studied in Caucasian populations of 18 rock lizard species of the genus Darevskia. Four subfamilies (Caucasian Lacerta satellites (CLsat)I-IV) were identified, which shared 70%-75% sequence similarity. The distribution of CLsat subfamilies among the species was studied. All the species could be divided into at least 3 clades, depending on the content of CLsat subfamilies in each genome: "saxicola", "rudis", and "mixta" lizards. CLsatI was found in all studied species, but in very different quantities; the "saxicola" group contained this subfamily predominantly. The "rudis" group also contained CLsatIII, and the "mixta" group carried considerable amounts of CLsatII. The highest concentrations of CLsatI and CLsatII were detected in 2 ground lizards--D. derjugini and D. praticola, respectively. D. parvula predominantly carried CLsatIII. CLsatIV was found only in the Crimean species D. lindholmi. The distribution patterns of satellite subfamilies show possible postglacial speciation within the genus Darevskia. A hybrid origin of species that possess 2 or 3 CLsat subfamilies and important clarifications to the systematics of the genus are proposed.

The 5S DNA sequences in Hordeum bogdanii and in the H. brevisubulatum complex, and the evolution and the geographic dispersal of the diploid Hordeum species (Triticeae: Poaceae)

The molecular diversity of 5S rDNA from the closely related Asiatic diploid species, Hordeum bogdanii and the H. brevisubulatum complex has been catalogued and analysed. As in previous studies in Hordeum , we found that the sequences are constrained in such an manner that unit classes can be defined. The long H1 unit class, known to occur in all Eurasian species, was frequently found in these 2 taxa. In addition, we identified a new unit class, called the short H3 to reflect the H genome found in these 2 taxa. Although the 2 taxa are very close morphologically, the variation in the long H1 DNA units is constrained to such a great degree that, in many cases, the accessions in a unit class from a single species are clustered. In H. bogdanii, the majority of the sequences are grouped in this manner, whereas in the H. brevisubulatum complex, the tendency to be constrained is lower in some but not all subspecies. These results support keeping H. brevisubulatum ssp. violaceum and ssp. iranicum as 1 species with the long H1 and short H1 unit classes, while retaining ssp. nevskianum and ssp. turkestanicum in the H. brevisubulatum complex. We have summarized our work on the presence/absence of the 10 unit classes found in all diploid species of Hordeum. A phylogenetic analysis, based strictly on the presence/absence of unit classes, indicated clearly that all the South American diploids and all the North American diploids possess long H2 and long Y2 unit classes and, except for H. californicum and H. pusillum, which contain long H1 in addition to the long H2 and long Y2 classes, are devoid of the long H1 unit class. This suggests that the gene gain/loss process from a common ancestor has been concomitant with intercontinental dispersal between the Old and the New Worlds.

Ancient differentiation of the H and I haplomes in diploid Hordeum species based on 5S rDNA

5S rDNA clones from 12 South American diploid Hordeum species containing the HH genome and 3 Eurasian diploid Hordeum species containing the II genome, including the cultivated barley Hordeum vulgare, were sequenced and their sequence diversity was analyzed. The 374 sequenced clones were assigned to "unit classes", which were further assigned to haplomes. Each haplome contained 2 unit classes. The naming of the unit classes reflected the haplomes, viz. both the long H1 and short I1 unit classes were identified with II genome diploids, and both the long H2 and long Y2 unit classes were recognized in South American HH genome diploids. Based upon an alignment of all sequences or alignments of representative sequences, we tested several evolutionary models, and then subjected the parameters of the models to a series of maximum likelihood (ML) analyses and various tests, including the molecular clock, and to a Bayesian evolutionary inference analysis using Markov chain Monte Carlo (MCMC). The best fitting model of nucleotide substitution was the HKY+G (Hasegawa, Kishino, Yano 1985 model with the Gamma distribution rates of nucleotide substitutions). Results from both ML and MCMC imply that the long H1 and short I unit classes found in the II genome diploids diverged from each other at the same rate as the long H2 and long Y2 unit classes found in the HH genome diploids. The divergence among the unit classes, estimated to be circa 7 million years, suggests that the genus Hordeum may be a paleopolyploid.

The utility of the nontranscribed spacer of 5S rDNA units grouped into unit classes assigned to haplomes - a test on cultivated wheat and wheat progenitors

Data is presented on the evolutionary dynamics of non-transcribed spacers (NTSs) of 5S rRNA genes in some diploid and polyploid Triticum and Aegilops species. FISH experiments with probes representing different unit classes revealed presence and (or) absence of these sequences in genomes or separate chromosomes of the species. Among the three diploid species only Aegilops speltoides has all of the different unit classes in ribosomal clusters as detected by the probes. Triticum urartu does not have the long D1 signals and Aegilops tauschii does not have the long A1 signals. Both polyploids possess all types of sequences, but because of genome rearrangements after polyploidization there is significant repatterning of single different rDNA unit classes in chromosomal positions when compared with those in diploid progenitors. Additional refined work is needed to ascertain if the sequences in the polyploids are mixed or are located in mini clusters in close proximity to each other. Mantel tests for association between the presence of the FISH signals of the A, B, and D genomes together and separately with the unit class data of the material, i.e., the probes used in FISH, indicated that all signals were associated with their respective probe material, but that there was no association of the unit classes found and the signals to each haplome. All combinations of the partial Mantel tests, e.g., between the A and B haplomes while controlling the effect of the all probes signals, with correlations ranging from 0.48 to 0.79 were all significant. Principal coordinate analysis showed that the signals of most unit class specific probes were more or less equally distant except for the long (S1 and short G1 signals, which were not different, and that the short A1 signals were closely related to the former two, whereas the signals of the long G1 were even less related.

Molecular diversity of the 5S rDNA units in theElymus dahuricuscomplex (Poaceae: Triticeae) supports the genomic constitution ofSt,Y, andHhaplomes

The phylogenetic analysis of 118 5S rRNA gene sequences cloned from members of the Elymus dahuricus complex containing the St, Y, and H haplomes, and of several related species containing at least one of these three haplomes, is reported. Differences in sequence pattern, primarily within the nontranscribed spacer, enabled the identification of six putative orthologous groups that we refer to as unit classes. In previous publications, we have been able to assign unit classes to haplomes. In addition to four unit classes previously identified in other genera, namely the long H1, long S1, long P1, and long {Y1, here we document two new unit classes called the long S2 and long W1. Most sequences of the E. dahuricus complex and related tetraploid species are classified as long S1 and assigned to the St haplome. Both long S1 and long S2 unit classes were identified in the diploid Pseudoroegneria spicata (Pursh) Á. Löve with the St haplome. The long S2 unit class was also identified in the hexaploid Elymus scabrus (R. Br.) Á. Löve with the St,Y,and W haplomes. The long P1 was known from the diploid Agropyron cristatum Gaertn. with the P haplome, and the long W1 was determined in Australopyrum retrofractum (Vickery) Á. Löve, known to contain the W haplome, but was not yet detected in E. scabrus, a hexaploid species with W being one of the three haplomes. The long H1 reported earlier from Hordeum was identified in several clones of the E. dahuricus complex. As previously reported, the long {Y1 unit class was found to be rare overall, but we identified it in a few clones of Elymus drobovii and in the E. dahuricus complex.Key words: 5S rDNA, unit classes, haplomes, concerted evolution.