Search

Fluorescence in situ hybridization (FISH) is a widely used method to localize DNA sequences on chromosomes. Out of the many uses, FISH facilitates construction of physical maps by ordering contigs of large-insert DNA clones, typically bacterial artificial chromosome (BAC) and establishing their orientation. This is important in genomic regions with low recombination frequency where genetic maps suffer from poor resolution. While BAC clones can be mapped directly by FISH in plants with small genomes, excess of repetitive DNA hampers this application in species with large genomes. Mapping single-copy sequences such as complementary DNA (cDNA) is an attractive alternative. Unfortunately, localization of single-copy sequences shorter than 10 kb remains a challenging task in plants. Here, we present a highly efficient FISH technique that enables unambiguous localization of single copy genes. We demonstrated its utility by mapping 13 out of 15 full-length cDNAs of variable length (2,127-3,400 bp), which were genetically defined to centromeric and pericentromeric regions of barley chromosome 7H. We showed that a region of 1.2 cM (0.7 %) on genetic map represented more than 40 % of the physical length of the chromosome. Surprisingly, all cDNA probes occasionally revealed hybridization signals on other chromosomes, indicating the presence of partially homologous sequences. We confirmed the order of 10 cDNA clones and suggested a different position for three cDNAs as compared to published genetic order. These results underline the need for alternative approaches such as FISH, which can resolve the order of markers in genomic regions where genetic mapping fails.

Intergeneric hybrids of fescues (Festuca spp.) and ryegrasses (Lolium spp.) are unique for the ability of their chromosomes to pair essentially freely in meiotic metaphase I (MI). At the same time, their chromosomes can be readily recognized by genomic in-situ hybridization (GISH). Past genome-wide observations suggested that this homoeologous pairing was not completely random. In this study we extend the analysis to all seven individual chromosomes of F. pratensis introgressed into autotetraploid L. multiflorum and show that for any F. pratensis chromosome the choice of an MI pairing partner depends on the identity of the remaining chromosomes present in the quadruplet. In monosomic introgressions, the choice of a homologous or homoeologous partner was completely random; in disomics there was a slight preference for homologous pairing. Pairing preference was similar for each chromosome, suggesting that pairing affinity of all chromosomes is essentially the same and no structural rearrangements differentiate the two genera. Homoeologous crossover rates for individual chromosomes were similar and they were consistently lower than expected on the basis of the MI pairing. High homoeologous MI pairing in these hybrids may be due to a very permissive system of chromosome pairing control that overlooks differences between the parental chromosomes. Given the ease of genome discrimination by GISH in the Lolium-Festuca hybrids, the differences in repetitive DNA sequences must be substantial. On the other hand, it appears just as likely that while the DNA repeats diverged markedly during evolution, the sequences involved in chromosome pairing have been conserved enough to facilitate regular pairing partner recognition and crossing-over.