ISSN 10227954, Russian Journal of Genetics, 2010, Vol. 46, No. 9, pp. 1036–1038. © Pleiades Publishing, Inc., 2010.
Original Russian Text © V.N. Babenko, R.B. Aitnazarov, F.A. Goncharov, I.F. Zhimulev, 2010, published in Genetika, 2010, Vol. 46, No. 9, pp. 1168–1170.
Several levels are known for the organization of
eukaryotic chromatin . There are both global and
local levels, which are regulated by an intricate system
involving both DNA elements and ribonucleoprotein
complexes and additionally utilizing histone modifi
cation, DNA methylation, and other epigenetic fac
tors. It is commonly accepted that the structural and
spatial organization of chromatin affects the regula
tion of gene expression .
A domain organization with respect to gene density
has been established for higher eukaryotic chromo
somes . Housekeeping genes often cluster in regions
with a high gene density, while tissuespecific genes
often occur in regions with a low gene density [1, 2].
We have analyzed the gene density distribution and
have identified 218 euchromatic genome regions with
a barbell structure. A barbell region has been defined
as a region consisting of three consecutive regions, the
gene density being distinctly lower in the central
region (barbell handle, BH) and higher at the flanks
In this work, we studied the alternative splicing
intensity for BH regions and their flanks. While the
alternative splicing intensity was much the same in BH
regions and their flanks, the distribution of the number
of isoforms per gene in the BH and flanking regions
proved to significantly differ depending on the gene
MATERIALS AND METHODS
Gene annotations were taken from the
genome version 5.12 (www.flybase.org).
We studied a total of 14058 genes from euchromatic
regions of chromosomes X, 2L, 2R, 3L, and 3R.
Intron density (number of introns per gene) was com
pared for genes from BH and flanking regions by
test. The number of isoforms was compared
for BH and flanking regions with the use of
fourcell tables. Approximation of the intron density
distribution with the
distribution was performed
using STATISTICA v. 6 package (StatSoft Inc.).
Alternative Splicing Landscape
V. N. Babenko
, R. B. Aitnazarov
, F. A. Goncharov
, and I. F. Zhimulev
Institute of Cytology and Genetics, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 630090 Russia;
Institute of Chemical Biology and Fundamental Medicine, Siberian Branch,
Russian Academy of Sciences, Novosibirsk, 630090 Russia
Received January 28, 2010
—Alternative splicing (AS) intensity (isoform number per gene) was studied as dependent on the
gene size for various regions of the
genome. The AS intensity of long transcripts from
regions with a low gene density proved to be significantly higher than for regions with a high gene density. An
opposite pattern was observed for small genes. The intron density distribution was approximated using the
distributions for regions with a high or low gene density. Statistical comparisons of the
firmed a lower coefficient
for regions with a low gene density (i.e., the average intron density was higher).
Based on these data, relaxed evolution of the exon–intron structure was assumed for regions with a low gene
Twosample Ftest for the variance of the intron density in
genes larger than 13 kb in size from different regions
Number of introns per transcript
Number of transcripts