Biomedical Engineering Reference
In-Depth Information
various shufflings of the original sequence's bases to new positions, we showed that
nearest-neighbor dinucleotide pair frequency conservation alone was sufficient to
account for nearly all of the average T m melting behavior of the real sequences. This is
in agreement with the accepted dependence observed in experimental DNA melting.
However, rather than the single classic Marmur-Doty linear relationship often cited in
the literature (148,151) to exist for all DNAs, we demonstrated that different relation-
ships existed, one for exon sequences (coding regions) and another for intron sequences
(noncoding regions). These two sequence classes were statistically different by 6.6% in
their slopes. These two dependencies could be explained by their slightly different
biases in nearest-neighbor dinucleotide pair frequency distributions. This study helped
establish the widespread utility of MELTSIM to accurately simulate the sequence
dependence of DNA melting behavior.
In 1996, Saccharomyces cerevisiae (baker's yeast) was the first eukaryotic genome
(12,067,277 bp) to be fully sequenced (153). We have shown that MELTSIM accurately
simulates the experimentally determined T m value of this entire genome (154). An
important feature of MELTSIM allows the simulation and display of the positional melt-
ing behavior along any DNA sequence. We have used this positional mapping feature to
demonstrate melting behavior along the 16 individual yeast chromosomes (154,155). As
Figure 1.47 demonstrates, the positional mapping feature of MELTSIM can also reveal
the positions of block DNA melting exhibited, for example, within a single gene: in this
case a diphosphatase enzyme from D. discoideum . There is a close correspondence
between the positions of introns and UTR regions of the diphosphatase gene and low-
melting temperature block behavior. Exon positions correspond to the high-melting
block regions of the diphosphatase gene. These block melting data are consistent with
the known base composition and dinucleotide composition of exons and introns and
their correlations with DNA melting temperature (121,151,152). This exon vs. intron dis-
crimination within the diphosphatase gene is just one of a number of such cases we
identified within sequences of the (A
T)% base -rich D. discoideum (156,157) and
Schistosoma mansoni (158) genomic DNAs based upon their differential positional melt-
ing behaviors. Such behavior is likely to be a general property of (A
T)% base-rich
eukaryotic genomes.
We believe that the ability to simulate the melting behavior of long DNA sequences
using MELTSIM has important potential uses in the understanding of DNA stability for
sequences of considerable length incorporated into biosensors. More complex biosensor
systems will eventually incorporate longer DNA sequences, lengths for which T m calcu-
lation programs designed for short oligonucleotide lengths are not appropriately para-
meterized (159). Given that the thermodynamic equilibrium block melting size for DNA
can be as little as 100 bp, we believe that in the future the understanding of complex
melting will be necessary to successfully deploy the intelligent properties of long DNA
sequences within biosensors. As a practical example of this importance, we describe the
current situation with microarrays. Nucleic acid-based microarrays are recognized as
increasingly important measurement tools in the basic research, biotechnology, and
pharmaceutical industries for characterization of the complex mRNA transcriptional
status of tens of thousands of genes within cells and tissues (160). It is perhaps not
widely appreciated that these microarrays are fundamentally biosensors. They rely typ-
ically upon single-stranded nucleic acid probe sequences being bound to small spotted
areas [(~70
m) 2 area], one sequence/spot, upon a suitable substrate. The spots are
grouped into large-array formats (tens of thousands of spots) that allow for simultane-
ous detection and quantitation of huge numbers of gene product mRNA sequences.
Following total mRNA isolation from a biological sample of interest, fluorescent chro-
mophore-labeled cDNA copies of these gene products are typically placed upon the
Search WWH ::




Custom Search