146
Alternatively Spliced Genes
of
searching
for
small
exons
in
the
vast
sea
of
introns.
This
process
of
‘‘exon recognition’’ or ‘‘exon deFnition’’ is
particularly remarkable because sequences
at mammalian splice sites are so highly
degenerate (see ±ig. 2). A large number
of sequence elements with similarity to
authentic splice sites (pseudosplice sites)
can be found in both intronic and exonic
regions). ±urther complicating the issue,
some of these elements act as
cryptic splice
sites
that are only used by the splicing
machinery
when
the
authentic
splice
sites are altered by mutations. Therefore,
speciFc recognition of authentic splice
sites and correct pairing of corresponding
5
0
and 3
0
splice sites is a central issue for
both constitutive splicing and alternative
splicing regulation. The intrinsic sequence
degeneracy
of
mammalian
splice
sites
determines
that
alternative
splicing
is
a rule rather than an exception during
mammalian gene expression.
In addition to splice sites, sequence ele-
ments in both intronic and exonic regions
modulate alternative splicing. Such reg-
ulatory elements can either enhance or
suppress splicing, and are hence named
exonic splicing enhancers (ESEs) and in-
tronic splicing enhancers (ISEs), or exonic
splicing silencers (ESSs) and intronic splic-
ing silencers (ISSs).
A number of ESE motifs have been iden-
tiFed using biochemical systematic evolu-
tion of ligands by exponential enrichment
(SELEX) or bioinformatical approaches.
A/G-rich (also called purine-rich) and A/C-
rich elements are among ESE motifs char-
acterized by biochemical studies. Proteins
containing SR domains play a major role in
recognizing A/G-rich ESEs and recruiting
other spliceosomal components (includ-
ing snRNPs and other protein factors),
thereby promoting the usage of neighbor-
ing splice sites. A cold-box protein, YB-1,
has been shown to enhance splicing by
interacting with an A/C-rich ESE. Another
ESEinH
IV
-1tev
-
spe
c
iF
cexonin
te
ra
c
t
s
with hnNRP H and SR protein SC35 to
enhance splicing. The SELEX method has
been used to identify preferred binding
sequences for individual SR proteins, and
optimal binding sites for individual SR pro-
teins are degenerate. ESE prediction pro-
grams based on SELEX and computational
analysis of human genes have been de-
programs are useful in predicting alterna-
tive splicing patterns of natural pre-mRNA
substrates in cells.
ISE
elements
have
been
studied
in
a number of genes, including c-src,
β
-
tropomyosin, calcitonin/calcitonin gene-
related peptide gene (CGRP), Fbronectin,
nonmuscle myosin heavy chain, cardiac
troponin T (cTNT), ±G±R-2,
α
2 subunit
of glycine receptor, and other genes. Such
ISEs may contain sequences similar to
5
0
splice site, U-rich element adjacent to
the regulated 5
0
splice site, UGCAUG el-
ement, (UCAUY)3-containing sequences,
or CUG-containing motif.
A number of ESSs have been char-
acterized in different genes such as
β
-
tropomyosin, CD44, and viral genes in-
cluding HIV Tat, bovine papillomavirus
type-1, and Rous sarcoma virus. They do
not share any obvious sequence motifs.
Their activities are often associated with
interactions with proteins of the hnRNP
family, including hnRNPA1, hnRNP H,
and hnRNP ±.
A
variety
of
ISSs
have
been
ana-
lyzed in alternatively spliced genes in-
cluding
hnRNP
A1,
Fbroblast
growth
factor
receptor
2
(±G±R2),
caspase
2,
GABA
A
R
γ
2(
γ
-aminobutyric acid receptor
typeA
γ
2subunit), NMDA R1 (
N
-methyl-
D-aspartate receptor R1 subunit;), clathrin
light chain B, and HIV tat.
Many of
previous page 146 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online next page 148 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online Home Toggle text on/off