Structural and functional aspects of grass genome organization

Alexander Pavlovich Tikhonov, Purdue University

Abstract

Studies of genome organization are important for understanding genome functions and evolution, and for developing strategies for its manipulation. To analyze the gene and repetitive DNA composition, evolution and putative chromatin domain organization of large genomic segments, we have sequenced 217 kb and 78 kb around the adh1 genes of maize and sorghum, respectively. Comparative analysis of the regions revealed conservation of gene composition, their order and direction of transcription. All maize candidate genes have homologs in orthologous positions and predicted directions of transcription in sorghum. However, two regions carrying at least three additional candidate genes in sorghum were deleted from maize. Average gene density is one gene per 25 kb and 5.5 kb for the maize and sorghum regions, respectively. A gene-dense maize region was found with <10 kb>/gene. About 92% of the intergenic sequence in the maize region is composed of LTR-retrotransposons; that form six blocks of 22 nested LTR-retrotransposons. These blocks are 14 kb to 70 kb in size. A massive invasion of LTR-retrotransposons into the maize adh1-F region occurred within the last 3–6 million years and contributed to the expansion of the region from 59 to 225 kb, which corresponds to >3-fold increase in genome size. No retrotransposons were identified in the sorghum region. Numerous miniature inverted-repeat transposable elements (MITEs) were uncovered, but no MITEs were detected within the 166-kb occupied by LTR-retrotransposons. To compare the putative higher-order structure of the regions, Matrix Attachment Regions (MARs) were mapped. Structural colinearity of putative chromatin domains and segregation of genes into comparable structural units was observed in two unlinked adh1 and sh2/a1 regions that suggests common principles of higher-order organization of plant genomes. Over 40 MARs were identified from the grass species. MITEs often co-localized with MARs. In vitro binding assays revealed that MITEs often have high affinities for the nuclear matrix. Locations of the MARs relatively to repetitive and gene sequences were determined. Maize MARs were localized between genes and retroelement blocks, sequestering most repetitive DNAs and genes into separate domains. Sequence analysis and computer prediction of MARs is presented.

Degree

Ph.D.

Advisors

Avramova, Purdue University.

Subject Area

Molecular biology|Genetics

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS