Relationships in the genus Hordeum: Giemsa C-banded karyotypes

On the allopolyploid origin and genome structure of the closely related species Hordeum secalinum and Hordeum capense inferred by molecular karyotyping

BACKGROUND AND AIMS

To provide additional information to the many phylogenetic analyses conducted within Hordeum , here the origin and interspecific affinities of the allotetraploids Hordeum secalinum and Hordeum capense were analysed by molecular karyotyping.

METHODS

Karyotypes were determined using genomic in situ hybridization (GISH) to distinguish the sub-genomes and , plus fluorescence in situ hybridization (FISH)/non-denaturing (ND)-FISH to determine the distribution of ten tandem repetitive DNA sequences and thus provide chromosome markers.

KEY RESULTS

Each chromosome pair in the six accessions analysed was identified, allowing the establishment of homologous and putative homeologous relationships. The low-level polymorphism observed among the H. secalinum accessions contrasted with the divergence recorded for the sub-genome of the H. capense accessions. Although accession H335 carries an intergenomic translocation, its chromosome structure was indistinguishable from that of H. secalinum .

CONCLUSION

Hordeum secalinum and H. capense accession H335 share a hybrid origin involving Hordeum marinum subsp. gussoneanum as the genome donor and an unidentified genome progenitor. Hordeum capense accession BCC2062 either diverged, with remodelling of the sub-genome, or its genome was donated by a now extinct ancestor. A scheme of probable evolution shows the intricate pattern of relationships among the Hordeum species carrying the genome (including all H. marinum taxa and the hexaploid Hordeum brachyantherum ).

Phylogenetic analysis of two single-copy nuclear genes revealed origin and complex relationships of polyploid species of Hordeum in Triticeae (Poaceae)

Two single-copy nuclear genes, the second largest subunit of RNA polymerase II (RPB2) and thioredoxin-like gene (HTL), were used to explore the phylogeny and origin of polyploid species in Hordeum. Our results were partly in accord with previous studies, but disclosed additional complexity. Both RPB2 and HTL trees confirmed the presence of Xa genome in H. capense and H. secalinum, and that H. depressum originated from H. californicum together with other American diploids, either H. intercedens or H. pusillum. American diploids solely contributed to the origin of H. depressum. The Asian diploids, either H. bogdanii or H. brevisubulatum, contributed to the formation of American polyploids except H. depressum. RPB2 and HTL sequences showed that H. roshevitzii did not contribute to the origin of American tetraploids. Our data showed a close relationship between the hexaploids H. procerum and H. parodii and the tetraploids H. brachyantherum, H. fuegianum, H. guatemalense, H. jubatum, and H. tetraploidum. The involvement of the diploid H. pusillum and the tetraploid H. jubatum in the formation of H. arizonicum was also indicated in the HTL phylogeny. Our results suggested a possible gene introgression of W- and P-genome species into the tetraploid H. jubatum and the hexaploid H. procerum.

Chromosomal organization of repetitive DNAs in Hordeum bogdanii and H. brevisubulatum (Poaceae)

Molecular karyotypes of Hordeum bogdanii Wilensky, 1918 (2n = 14), and Hordeum brevisubulatum Link, 1844 ssp. brevisubulatum (2n = 28), were characterized by physical mapping of several repetitive sequences. A total of 18 repeats, including all possible di- or trinucleotide SSR (simple sequence repeat) motifs and satellite DNAs, such as pAs1, 5S rDNA, 45S rDNA, and pSc119.2, were used as probes for fluorescence in situ hybridization on root-tip metaphase chromosomes. Except for the SSR motifs AG, AT and GC, all the repeats we examined produced detectable hybridization signals on chromosomes of both species. A detailed molecular karyotype of the I genome of Hordeum bogdanii is described for the first time, and each repetitive sequence is physically mapped. A high degree of chromosome variation, including aneuploidy and structural changes, was observed in Hordeum brevisubulatum. Although the distribution of repeats in the chromosomes of Hordeum brevisubulatum is different from that of Hordeum bogdanii, similar patterns between the two species imply that the autopolyploid origin of Hordeum brevisubulatum is from a Hordeum species with an I genome. A comparison of the I genome and the other Hordeum genomes, H, Xa and Xu, shows that colocalization of motifs AAC, ACT and CAT and colocalization of motifs AAG and AGG are characteristic of the I genome. In addition, we discuss the evolutionary significance of repeats in the genome during genome differentiation.

Molecular phylogeny of diploid Hordeum species and incongruence between chloroplast and nuclear datasets

The phylogeny of diploid Hordeum species has been studied using both chloroplast and nuclear gene sequences. However, the studies of different nuclear datasets of Hordeum species often arrived at similar conclusions, whereas the studies of different chloroplast DNA data generally resulted in inconsistent conclusions. Although the monophyly of the genus is well supported by both morphological and molecular data, the intrageneric phylogeny is still a matter of controversy. To better understand the evolutionary history of Hordeum species, two chloroplast gene loci (trnD-trnT intergenic spacer and rps16 gene) and one nuclear marker (thioreoxin-like gene (HTL)) were used to explore the phylogeny of Hordeum species. Two obviously different types of trnD-trnT sequences were observed, with an approximately 210 base pair difference between these two types: one for American species, another for Eurasian species. The trnD-trnT data generally separated the diploid Hordeum species into Eurasian and American clades, with the exception of Hordeum marinum subsp. gussoneanum. The rps16 data also grouped most American species together and suggested that Hordeum flexuosum has a different plastid type from the remaining American species. The nuclear gene HTL data clearly divided Hordeum species into two clades: the Xu+H genome clade and the Xa+I genome clade. Within clades, H genome species were well separated from the Xu species, and the I genome species were well separated from the Xa genome species. The incongruence between chloroplast and nuclear datasets was found and discussed.

Genome size in natural and synthetic autopolyploids and in a natural segmental allopolyploid of several Triticeae species

Nuclear DNA amount (1C) was determined by flow cytometry in the autotetraploid cytotype of Hordeum bulbosum, in the cytologically diploidized autotetraploid cytotypes of Elymus elongatus, Hordeum murinum subsp. murinum and Hordeum murinum subsp. leporinum, in Hordeum marinum subsp. gussoneanum, in their progenitor diploid cytotypes, and in a newly synthesized autotetraploid line of E. elongatus. Several lines collected from different regions of the distribution area of every taxon, each represented by a number of plants, were analyzed in each taxon. The intracytotype variation in nuclear DNA amount of every diploid and autotetraploid cytotype was very small, indicating that no significant changes have occurred in DNA amount either after speciation or after autopolyploid formation. The autotetraploid cytotypes of H. bulbosum and the cytologically diploidized H. marinum subsp. gussoneanum had the expected additive amount of their diploid cytotypes. On the other hand, the cytologically diploidized autotetraploid cytotypes of E. elongatus and H. murinum subsp. murinum and H. murinum subsp. leporinum had considerably less nuclear DNA (10%-23%) than the expected additive value. Also, the newly synthesized autotetraploid line of E. elongatus showed similar reduction in DNA as its natural counterpart, indicating that the reduction in genome size occurred in the natural cytotype during autopolyploidization. It is suggested that the diploid-like meiotic behavior of these cytologically dipolidized autotetraploids is caused by the instantaneous elimination of a large number of DNA sequences, different sequences from different homologous pairs, leading to differentiation of the constituent genomes. The eliminated sequences are likely to include those that participate in homologous recognition and initiation of meiotic pairing. A gene system determining exclusive bivalent pairing by utilizing the differentiation between the two groups of homologues has been presumably superimposed on the DNA reduction process.