Intraspecific and interspecific polyploidy of Brazilian species of the genus Inga (Leguminosae: Mimosoideae).

We investigated the karyotypes of 13 species of six sections of the genus Inga (Leguminosae-Mimosoideae) from Brazil. We used conventional Giemsa staining to identify numerical chromosomal variations and looked for karyotypic evolutionary patterns. The karyotypes generally had small chromosomes, varying from metacentric to submetacentric, with a basic number x=13. Nine of the species showed 2n=2x=26 (I. thibaudiana, I. cayennensis, I. ingoides, I. edulis, I. vera, I. subnuda, I. striata, I. bollandii, and Inga sp), while 2n=4x=52 was seen in a population of Inga cylindrical and of I. capitata, and in five populations of I. laurina. Additionally, 2n=8x=104 was observed in a population of I. cayennensis. Eight of these counts were new, while the counts of 2n=52 for I. laurina and 2n=26 for I. marginata, I. vera, I. subnuda, and I. edulis confirmed previous studies. We did not find cytological stability among the sections studied, with occurrence of significant intra- and inter-specific numerical variations. We conclude that polyploidy has played a significant role in karyotypic evolution in this group and that it occurred independently in several sections of the genus.

Linked 5S and 45S rDNA sites are highly conserved through the subfamily Aurantioideae (Rutaceae).

Sites of 5S and 45S rDNA are more commonly located on different chromosomes of most angiosperms. Previous investigations have shown that in the subfamily Aurantioideae these sites may appear closely linked (adjacent sites), as in Poncirustrifoliata, or completely isolated, as in some species of Citrus. In the present work, the distribution of rDNA sites was investigated in representatives of 9 genera of Aurantioideae by FISH and CMA banding, aiming to understand the evolution of adjacent sites in the subfamily. A total of 57 rDNA sites were observed, 40 of them being adjacent to each other. All adjacent sites displayed the 45S rDNA array more terminally located. Assuming that the linked 5S-45S rDNA arrangement was the ancestral condition in Aurantioideae, the isolated rDNA sites observed in Clausena excavata,Bergera koenigii, and Fortunella obovata, as well as the complete linkage loss in Citrus maxima and C. medica indicates that unlinked sites arose independently several times in the evolution of the group. The linkage loss may be due to independent dispersion of one or both rDNA sequence families followed by deletion of the corresponding array in the adjacent site. The possible mechanisms behind these events and their occurrence in other groups are discussed.

Distribution of 5S and 45S rDNA sites in plants with holokinetic chromosomes and the "chromosome field" hypothesis.

Secondary constrictions or 45S rDNA sites are commonly reported to be located mainly in the terminal regions of the chromosomes. This distribution has been assumed to be related to the existence of a "chromosome field" lying between the centromere and the telomere, an area in which certain cytogenetic events may predominantly occur. If this hypothesis is true this distribution should not be observed in holokinetic chromosomes, as they do not have a localized centromere. In order to evaluate this hypothesis, a comparative study was made of the distributions of 5S and 45S rDNA sites using fluorescence in situ hybridization in representatives of the genera Eleocharis, Diplacrum, Fimbristylis, Kyllinga and Rhynchospora, all of which belong to the family Cyperaceae. The numbers of sites per diploid chromosome complement varied from 2 to ∼10 for 5S rDNA, and from 2 to ∼45 for 45S rDNA. All of the 11 species analyzed had terminally located 45S rDNA sites on the chromosomes whereas the 5S rDNA sites also generally had terminal distributions, except for the Rhynchospora species, where their position was almost always interstitial. These results, together with other previously published data, suggest that the variation in the number and position of the rDNA sites in species with holokinetic chromosomes is non-random and similar to that reported for species with monocentric chromosomes. Therefore, the predominant terminal position of the 45S rDNA sites does not appear to be influenced by the centromere-telomere polarization as suggested by the "chromosome field" hypothesis. Additionally, the hybridization of 5S and 45S rDNA sites provides interesting markers to distinguish several chromosomes on the rather symmetrical karyotypes of Cyperaceae.

The meaning of DAPI bands observed after C-banding and FISH procedures.

Under specific technical conditions chromosome staining with 4',6-diamidino-2-phenylindole (DAPI) permits characterization of heterochromatic regions as AT-rich (DAPI(+)) or AT-poor (DAPI(-)), especially when the chromosomes are counterstained with chromomycin A(3) (CMA), which preferentially binds to GC-rich DNA. DAPI(+) bands also often have been observed after C-banding or FISH. In these cases, however, it is not clear whether only AT-rich regions stain positively with DAPI or other heterochromatins with different base compositions also are stained. We evaluated the meaning of DAPI bands observed after C-banding and FISH using three plant species bearing different types of heterochromatin: DAPI(+)/CMA(-), DAP(-)/CMA(+) and DAPI(0)/CMA(0) (neutral bands). Additional tests were performed using propidium iodide, a fluorochrome without preferential affinity for AT or GC. Our results indicate that AT-rich heterochromatin stains as DAPI(+) bands after C-banding or FISH, but other kinds of heterochromatin also may be stained by DAPI.

Heterochromatin banding patterns in Rutaceae-Aurantioideae--a case of parallel chromosomal evolution.

The heterochromatin banding patterns in the karyotypes of 17 species belonging to 15 genera of Rutaceae subfamily Aurantioideae (= Citroideae) were analyzed with the fluorochromes chromomycin (CMA) and 4'-6-diamidino-2-phenylindole-2HCl (DAPI). All species were diploids, except one tetraploid (Clausena excavata) and two hexaploids [Glycosmis parviflora agg. (aggregate) and G. pentaphylla agg.]. There are only CMA/DAPI bands, including those associated with the nucleolus. Using recent cpDNA (chloroplast DNA) sequence data as a phylogenetic background, it becomes evident that generally more basal genera with rather plesiomorphic traits in their morphology, anatomy, and phytochemistry exhibit very small amounts of heterochromatin (e.g., Glycosmis, Severinia, Swinglea), whereas relatively advanced genera from different clades with more apomorphic characters display numerous large CMA bands (e.g., Merrillia, Feroniella, Fortunella). Heterochromatin increase (from 0.7 to 13.7%) is interpreted as apomorphic. The bands are mostly located in the larger chromosomes and at telomeric regions of larger arms. However, one of the largest chromosome pair has been conserved throughout the subfamily with only very little heterochromatin. The heterochromatin-rich patterns observed in different clades of Aurantioideae appear quite similar, suggesting a kind of parallel chromosomal evolution. In respect to the current classification of the subfamily, it is proposed to divide Murraya s.l. (sensu lato) into Bergera and Murraya s.s. (sensu stricto) and to place the former near Clausena into Clauseneae s.s. and the latter together with Merrillia into Citreae s.l. The subtribes recognized within Clauseneae s.s. and Citreae s.l. appear heterogeneous and should be abandoned. On the other hand, the monophyletic nature of the core group of Citrinae, i.e., the Citrus clade with Eremocitrus, Microcitrus, Clymenia, Poncirus, Fortunella, and Citrus, is well supported.