Trends in Genetics
Volume 19, Issue 3, March 2003, Pages 119-124
Journal home page for Trends in Genetics

Genome Analysis
A significant fraction of conserved noncoding DNA in human and mouse consists of predicted matrix attachment regions

https://doi.org/10.1016/S0168-9525(03)00016-7Get rights and content

Abstract

Noncoding DNA in the human–mouse orthologous intergenic regions contains ‘islands’ of conserved sequences, the functions of which remain largely unknown. We hypothesized that some of these regions might be matrix–scaffold attachment regions, MARs (or S/MARs). MARs comprise one of the few classes of eukaryotic noncoding DNA with an experimentally characterized function, being involved in the attachment of chromatin to the nuclear matrix, chromatin remodeling and transcription regulation. To test our hypothesis, we analyzed the co-occurrence of predicted MARs with highly conserved noncoding DNA regions in human–mouse genomic alignments. We found that 11% of the conserved noncoding DNA consists of predicted MARs. Conversely, more than half of the predicted MARs co-occur with one or more independently identified conserved sequence blocks. An excess of conserved predicted MARs is seen in intergenic regions preceding 5′ ends of genes, suggesting that these MARs are primarily involved in transcriptional control.

Section snippets

MAR in the J-C intron of Igκ gene

MARs in the rabbit and mouse Igκ genes (MARIgκ) were identified experimentally next to the intronic enhancer, and in human MARIgκ was predicted on the basis of sequence conservation [12]. MARIgκ performs important biological functions: the mouse Igκ gene construct without MARIgκ exhibits low and erratic expression in transgenic mice [17] and somatic hypermutation is reduced in MARIgκ-knockout mice [18]. The dot matrix alignment (ftp://ncbi.nih.gov/pub/kondrashov/MAR/align.zip) revealed

Predicted MARs in human–mouse orthologous genomic contigs

As already mentioned, MARs are among the few functionally characterized long noncoding sequences in eukaryotic genomes. Because the functions of the human–mouse HITs are largely unknown, we sought to determine whether these regions overlapped with MARs. To this end, we predicted MARs using the ChrClass program (ftp://ncbi.nih.gov/pub/kondrashov/MAR/Chrclass.zip; [9]). The observed frequencies of predicted MARs per 1000 bases (1 kb) did not differ significantly between mouse and human

References (25)

  • G.B. Singh

    Mathematical model to predict regions of chromatin attachment to the nuclear matrix

    Nucleic Acids Res.

    (1997)
  • C.M. van Drunen

    A bipartite sequence element associated with matrix/scaffold attachment regions

    Nucleic Acids Res.

    (1999)
  • Cited by (0)

    View full text