DNA SEQUENCE ANALYSIS OF A GENE-DENSE REGION OF Zea Mays

Waksman Institute, Rutgers University, 190 Frelinghuysen Road, Piscataway, NJ 08854-8020 USA
http://mbclserver.rutgers.edu/~messing

Genetic maps of plant genomes have already shown the uneven distribution of gene sequences within chromosomes. Gene-dense regions frequently involve arrays of related genes. In Zea mays, the 22-kDa alpha-zein genes, encoding storage proteins, are clustered in one location in chromosome 4S and are a model for the study of a gene-dense region. We have constructed a high-resolution genetic and physical map. The 3.4-cM 22-kDa alpha-zein cluster region spans less than 250 kb, and includes two subclusters of zein genes, separated by a large (70kb) intervening spacer region. We have isolated cosmids and cDNAs containing 22-kDa zein sequences, and cosmids from the intervening spacer region using the RFLP marker php20725 from the inbred line BSSS53. We have sequenced two cosmids, one set of overlapping cosmids including 22-kDa zein-related sequences, and two overlapping cosmids from the intervening spacer. Over 200 kb of maize genomic DNA sequence has been analyzed with respect to spacing of genes and repetitive elements. Random sequencing of 130 22-kDa zein-specific cDNAs has provided a sampling of 22-kDa zein gene expression at the 1% level and revealed which of the 22-kDa zein genes produce stable transcripts. The majority of 22-kDa zein sequences are present in tandem arrays and contain many copies which do not produce stable transcripts. Although in a few cases this is due to truncations and insertions, in most cases the difference is correlated with the absence of full length open reading frames caused by mutations in the glutamine codon CAG or CAA that result in TAG and TAA stop codons. Interestingly, one gene in another inbred line, W22, has one in-frame stop codon, where its ortholog in BSSS53 has none and is expressed at normal levels. Therefore, we propose two gene pools, one acting as a reserve for the other. Weakly expressed genes could switch to normally expressed ones by gene conversion of orthologous sequences.