Article PDF - Landes Bioscience

bunkietalentedΤεχνίτη Νοημοσύνη και Ρομποτική

24 Νοε 2013 (πριν από 3 χρόνια και 8 μήνες)

182 εμφανίσεις

©2005 LANDES BIOSCIENCE. DO NOT DISTRIBUTE.
www.landesbioscience.com RNA Biology 53
[RNA Biology 2:2, 53-62; April/May/June 2005]; ©2005 Landes Bioscience
Andreas Werner
Institute for Cell and Molecular Biosciences; The Medical School; Framlington
Place; University of Newcastle; Newcastle upon Tyne UK
Correspondence to: Andreas Werner; Institute for Cell and Molecular Biosciences;
The Medical School; Framlington Place; University of Newcastle; Newcastle upon
Tyne NE2 4HH UK; Tel.: 44.191.222.6990; Fax: 44.191.222.7424; Email:
andreas.werner@ncl.ac.uk
Received 05/11/04; Accepted 05/19/04
Previously published as a RNA Biology E-publication
:
http://www.landesbioscience.com/journals/rnabiology/abstract.php?id=1852
KEY WORDS
antisense RNA, double stranded RNA, natural
antisense transcript, epigenetics, gene regulation
ACKNOWLEDGEMENTS
I would like to acknowledge the Novartis
Foundation for the generous support in organizing
the discussionmeeting“Natural antisense transcripts”;
the speakers Dr. Marjorie Matzke, Dr. Keith Brown,
Dr. Gordon Carmichael, Dr. Wolfgang Nellen and
Dr. Christopher Sanderson for their presentations
and all the participants of the meeting for discussion.
I would like to thank Mark Carlile and Gavin
McHaffie for the critical reading of the manuscript.
Review
Natural Antisense Transcripts
ABSTRACT
The sequencing of whole genomes and the subsequent annotation of cDNAs revealed
that about 20% of human and mouse genes overlap resulting in potential pairs of sense
and antisense transcripts. An increasing number of experimentally identified antisense
transcripts concur with this predication.
Characterization of overlapping transcripts in various species indicates that this form
of RNA-mediated gene regulation represents a widespread phenomenon. However, the
physiological relevance of natural antisense transcripts remains obscure. Genomic studies
suggest that duplex formation between sense and antisense is required for biological
function.
Antisense transcripts play an established role in imprinting and X-chromosome inacti-
vation and genomic rearrangements as observed in B and T leukocytes. Only a relatively
small percentage of the predicted antisense transcripts are related to these biological
phenomena that are also related to mono-allelic expression.
Consequently, there are at least two categories of natural antisense transcripts that
show significant differences with regard to their biological function as well as the potential
mechanisms involved.
INTRODUCTION
Antisense RNAs are linked to three important areas of epigenetic gene regulation. First,
and most important for this review, about 15–25% of mammalian genes overlap and
potentially interfere with each other’s expression.
1,2
Other eukaryotes such as Drosophila
3
or plants
4,5
show less but still very significant bi-directional transcription. The large
number of antisense transcripts emerged very recently and only a relatively small number
have been experimentally characterized. Second, double-stranded RNA has a distinct and
well described role in RNA interference and related mechanisms as well as in RNA editing.
6,7
Third, gene clusters that are only transcribed from one allele as observed in genomic
imprinting, X-chromosome inactivation or allelic exclusion in B- and T-cells express anti-
sense transcripts.
8
A discussion meeting on “Natural antisense transcripts” was recently held at the Novartis
Foundation in London. It assembled epigeneticists and scientists researching antisense
regulation of specific, physiologically relevant genes. The aim of the meeting was to integrate
expert knowledge towards a coherent picture and better understanding of natural antisense
transcription in vertebrates. In addition, a platform was generated for scientists working
on natural antisense RNAs (www.narna.ncl.ac.uk). This review summarizes these ongoing
efforts. Examples of antisense RNAs that influence the expression of their cognate sense
transcript will be discussed prominently. Genome wide approaches and computational
studies will add to a working hypothesis that integrates natural antisense transcripts into a
rapidly growing network of epigenetic gene regulation. Such model requires the consideration
of RNA interference (RNAi) and RNA editing. The review will specifically emphasize
aspects that could be relevant for the handling of natural antisense transcripts. In addition,
antisense RNAs play an important role in DNA methylation,
9
genomic imprinting,
10,11
and X-chromosome inactivation.
9,12,13
These subjects will be mentioned briefly, a detailed
description of all the indicated areas is beyond the scope of this article.

NATURAL ANTISENSE TRANSCRIPTS IN FOCUS
N
atural antisense transcripts overlap with
well-defined, physiologically relevant genes and
potentially downregulate the expression of the
complementary sense transcript (Fig. 1).
Computational screens have identified a large
number of overlapping genes in vertebrate
genomes; 22% of all genes overlap in humans,
2,14
about 15% in mice
1,15
and
Drosophila
3
and 6–9%
in plants,
4,5
respectively. Only few examples have
been characterized in experimental detail and the
physiological relevance of the large-scale antisense
transcription is speculative.
16-23
The widespread
occurrence and the apparent poor cross species
conservation of bi-directional transcription prompted
the rather nihilistic suggestion of “transcriptional
noise”.
24
Such noise it is argued would only be a problem if a cellular
system lacked the tools to deal with the “pollution”—and, with RNAi
and RNA editing such mechanisms seem to be in place. There are,
however, strong arguments that antisense transcripts, or at least a
substantial number of them, exhibit a biological function.
First, knock out animals with deleted antisense transcripts corrob-
orate the pivotal role of the cognate antisense RNAs. The best
documented examples include the imprinted antisense RNAs
Air
and
Tsix.
25,26
In both cases the antisense transcript is essential for
imprinted expression of the
Igf2r/Slc22a2/Slc22a3
cluster and random
X-chromosome inactivation, respectively. In addition, the knock-out
of the
frq
(frequency) antisense RNA was shown to affect the circadian
rhythm in
Neurospora crassa
.
27,28
This example also indicates that
gene regulation by antisense transcripts is well-conserved during
evolution.
Second, promoters of antisense transcripts contain transcription
factor binding sites that respond to physiological stimuli. The promoters
of fibroblast growth factor and the related antisense transcript share
Sp1 and Ets consensus sequences. The competence of the Ets site in
the antisense promoter was corroborated by electrophoretic mobility
shift assay.
29
On a much larger scale Cawley et al. mapped the tran-
scription factor binding sites for Sp1, cMyc and p53 on human
Chrs. 21 and 22.
24
O
nly 22% of these sites w
er
e found in the 5'
region of protein coding genes. In contrast, 36% of the binding sites
were found in the 3' region and correlated with the transcription of
non-coding RNAs. Retinoic acid affected the activity of both sense
and antisense related promoters often in a coordinate fashion.
24
Third, an increasing number of antisense transcripts have been
experimentally characteriz
ed and found to have distinct regulatory
roles related to the sense transcript.
21
This review describes selected
examples of characterized antisense transcripts without the ambition
of putting together a complete list (F
ig. 2; w
eb
-
based r
esources
provide updated figures: http://www.bioinfo.org.cn/NONCODE/
index.htm). Examples from mammals but also from non-mammalian
vertebrates and
Neurospora crassa
are included.
FGF-2/GFG
. Arguably one of the best-described example of a
bi
-
dir
ectionally transcribed gene encodes the fibr
oblast gr
o
wth
factor 2 (FGF-2) and its cognate antisense transcript
GFG-2
. The
sense-encoded protein induces growth, differentiation and anti-
apoptotic behavior; over expression of FGF-2 is linked to tumor
progression and ectopic cell pr
oliferation.
30
The FGF
-
r
elated anti
-
sense transcript was first identified in
Xenopus
oocytes. Interestingly,
Kimelman and Kirschner found that the sense and the antisense
transcript were edited in the overlapping region suggesting that the
two transcripts hybridized.
30-32
Subsequently the antisense transcript
was cloned from various species including chick, rat and human.
16,33,34
Sense and antisense transcripts overlap tail to tail in the 3' non-coding
region, only the
Xenopus
antisense transcript reaches into the open
reading frame of the sense transcript. Interestingly, the antisense
transcript encodes a conserved protein belonging to the family of
nucleoside phosphohydrolases (NUDT6). The expression of the
protein could be demonstrated by immunocytochemistry and Western
blot in liver and pituitary cells.
35,36
The function of the protein is
currently unknown. Studies using a pituitary tumor-derived cell line
(GH4) transfected with GFG-encoding cDNA constructs suggested
an antiproliferative effect of GFG protein independent of FGF-2
expression. Partly conflicting findings regarding the intracellular
location of the transgene product in C6 glioma cells could reflect cell
specific variations.
37
A putative regulatory impact of the antisense
transcript via RNA-RNA hybridization was originally suggested and
the fact that sense and antisense transcripts are inversely expressed in
various tissues support this hypothesis.
16,38
In addition, under patho-
physiological conditions aberrant ratios of sense/antisense transcrip-
tion are observed. Generally, reduced levels of the GFG-RNA favor
FGF
-
induced carcinogenic progression,
29
or
, in the case of
endometriosis ectopic implantation of endometric tissue.
30
HIF-1
α
.
The hypoxia inducible factor (HIF-1) is a heterodimeric
transcription factor that coordinates gene expression in response to
oxygen restriction. One component of the dimer, HIF-1
α
, is highly
r
egulated whereas HIF-1
β
is constitutiv
ely expressed. The gene
encoding HIF-1
α
is transcribed in both directions giving rise to an
antisense transcript that overlaps the 3' non-coding region of the
sense transcript in a tail to tail arrangement.
The antisense transcript
is unspliced, different splice variants have been reported of the sense
RNA.
The antisense transcript has also been demonstrated b
y
RT-PCR in mouse and rats.
39,40
According to the general belief, that
antisense RNA down regulates the sense transcript, the antisense
transcript antagoniz
es HIF
-
1 expr
ession in r
esponse to pr
olonged
hypoxia.
41
It is hypothesized that the antisense transcript exposes an
AU rich element in the 3’ untranslated region of the sense transcript,
destabilizes it and thus reduces its half-life. Another possibility suggests
that the antisense transcript inter
fer
es with the splicing of the sense
transcript resulting in the expression of a competing but non-func-
tional shor
ter splice v
ariant of HIF-1
α
.
42
The fact that HIF
-
1 induces
angiogenesis makes the regulation of the gene a highly relevant topic
in cancer r
esearch. Thrash-Binghof et al. reported that the HIF-1
54 RNA Biology 2005; Vol. 2 Issue 2
Natural Antisense Transcripts
Figure 1. Definition of natural antisense transcripts. All evidence thus far indicates that natural
antisense transcripts represent normally processed RNA polymerase II transcripts. The upper
panel shows a situation where the overlapping antisense RNA potentially hybridizes to the
spliced sense RNA. The scheme also holds for intronless genes. The non-overlapping transcripts
represented in the lower panel would only hybridize as heteronuclear RNAs.
www.landesbioscience.com
RNA
Biology 55
antisense transcript was strikingly over expressed in non-papillary
renal cancer but not in papillary carcinomas.
43
Msx1.
Msx1 is a homeobox binding transcription factor that is
particularly important during craniofacial development.
44
The
Msx1
gene comprises one intron with the homeobox encoding sequence
located in the second exon. The 3' end of the second exon also
constitutes the promoter region for an unspliced antisense transcript.
The
Msx1
antisense promoter is very well conserved between species.
The most striking feature about the
Msx1
sense/antisense transcripts
is their strict reciprocal expression at some stages of murine tooth
development. In situ hybridization on parallel sections of 16.5 day
embryos frontal incisors produced virtually mirror images with
Msx1
sense and antisense specific probes. The sense RNA was strongly
expressed in dental mesenchyme and the follicular sacs whereas the
antisense transcript was present in dental epithelium and weakly in
dental mesenchyme as well.
45
A negative regulatory effect of the
antisense RNA on Msx1 protein expression was demonstrated in the
osteoblast cell line MO6-G3.
45,46
The mechanism for this downreg-
ulation is not established, but could involve splicing or the degradation
of RNA hybrids.
α
-thyroid receptor/rev-ErbA
α
.
The
α
-thyroid receptor represents
another example of a transcription factor that is regulated by the
expression of an antisense transcript. Upon binding of thyroid
hormone (T
3
) the receptor activates a large number of genes generally
stimulating metabolism, growth and differentiation.
47
The mammalian
α
-thyroid receptor gene encodes two different splice forms, TR-
α
1
and TR-
α
2 with antagonistic properties. TR-
α
2 lacks hormone-
binding capacity and acts as a dominant repressor of TR-
α
1 and
other closely related receptors.
48
The structures of the TR-
α
1 and
TR-
α
2 encoding transcripts only differ at the 3' end due to an alter-
natively spliced last exon: Exon 9 of
TR-
α
1
encompasses 3400 bp
and includes stop codon and polyadenylation signals. In
TR-
α
2
exon 9 is 128 bp long and spliced to another exon (9A) downstream
of the
TR-
α
1
polyadenylation site.
49
Interestingly, the exon 9A over-
laps with another gene encoding an orphan nuclear receptor of the
same family denoted Rev-ErbA
α
. Rev-ErbA
α
is highly conserved in
all vertebrates, whereas the TR-
α
2/ Rev-ErbA
α
juxtaposition is only
found in mammals.
50,51
Both sense TR-
α
encoding transcripts and
the
Rev-ErbA
α
antisense transcript are expressed in a tissue specific
manner with
Rev-ErbA
α
correlating with the
TR-
α
1/TR-
α
2
ratio.
52,53
It is hypothesized that
Rev-ErbA
α
influences the processing of the
TR-
α
hnRNA towards formation of the
α
1
isoform.
52
The overlap
of
Rev-ErbA
α
and
TR-
α
2
is pivotal for this step as short RNAs com-
plementar
y to this r
egion block
TR
-
α
2
splicing.
54
The mechanism
triggered by the putative RNA-RNA hybrid is yet unclear. The fact
that mRNA stability is not significantly altered would argue for a
switch in
TR-
α
RNA processing and not degradation of the tran-
scripts b
y RNAi.
Myosin heavy chain (MHC).
A particularly interesting case of
antisense regulation is represented by genes encoding the two isoforms
of the my
osin heavy chain (MHC).
The car
diac muscle expr
esses an
α
and a
β
form with different ATPase activities. The phenotypic
composition of
α
-
and
β
-
MHC defines the contractile pr
oper
ties of
the cardiac muscle. The relative expression level of the two isoforms
is highly regulated during development and may also change under
pathophysiological conditions.
The two genes ar
e arranged head to
tail with the
α
isoform downstream of
β
-MHC
. Two independent
promoters drive the expression of the two isoforms. Luther et al.
reported an antisense transcript related to
β
-MHC
that was conserved
betw
een rat and human.
55,56
The antisense transcript was later sho
wn
Natural Antisense Transcripts
Figure 2. Schematic representation of bi-directionally transcribed genes with
physiological roles. Translated regions are represented in black, non-coding
regions are shown in grey. The size of exons and introns are not in scale.
Fgf-2
: Sense and antisense transcript overlap in the 3' untranslated region
of the
Fgf-2
transcript. The antisense transcript is translated. In
Xenopus
the
coding regions of sense and antisense transcript overlap.
Hif-1
α
: The
com
plementar
y transcripts overlap in the 3' untranslated region of
Hif
-
1
α
.
The antisense transcript lacks an open reading frame.
Msx1
: The promoter
region driving antisense expression encodes the homeobox of
Msx1
. The
antisense transcript is intronless and non-coding.
TR
α
1/TR
α
2/Rev-ErbA
α
:
The antisense transcript
Rev
-ErbA
α
overlaps with the last exon of the
TR
α
2
sense transcript. The black box indicates a hybrid between coding regions.
Both sense and antisense mRNA encode for receptors for T3 and an
unknown substrate, respectively. The expression of the antisense transcript
influences the splicing of the primary transcript towards generation of the
TR
α
1
isoform. Myosin heavy chain:
α
- and
β
-MHC
are represented without
detailing the intron exon structure. The bi-directional promoter between the
two isoforms is indicated. Transcription of the
α
-MHC
isoform results in
concomitant expression of
β
-MHC
antisense RNA. NaPi-II: The structure of
the Na/phosphate encoding sense transcript is well conserved between
zebrafish and mouse, the antisense transcript is not. There is a single com
mon
overlap at exon 10 between sense and antisense RNA in the two species.
The antisense in zebrafish is non-coding, the mouse transcript has short open
reading frame. In addition, the murine antisense transcript shares the
promoter, transcription start and the first 126 bases with Profilin III.
Frq
: Both
frq
sense and antisense transcripts are crucial in maintaining circadian
rhythm in
Neurospora crassa
. The non-coding antisense RNA contains a
short intron.
to initiate from the
α
-MHC
promoter that was active in both direc-
tions.
57
Consequently, stimulation of
α
-MHC
was paralleled by the
expression of
β
-MHC
antisense RNA and concomitant silencing of
the related isoform. The mechanism of silencing is currently
unknown; the methylation status of the
β
-MHC
promoter region
has not been addressed (yet). However, indirect evidence from trans-
genic mice expressing chloramphenicol acetyltransferase driven by
β
-MHC
promoter fragments indicated that the gene was not correctly
regulated. Specifically, the reporter gene was not down regulated in
the cardiac ventricle of transgenic animals after birth whereas
β
-MHC
normally is.
58
It was concluded that
β
-MHC
expression was
contr
olled b
y a do
wnstream element, that is, as we know now, most
likely the antisense RNA.
Sodium/phosphate cotransporter (NaPi-II).
The previous exam-
ples focused pr
edominatly on mammalian model system ho
w
ev
er
,
also lower vertebrates rely on gene regulation by natural antisense
transcripts. Our group has reported the identification of antisense
RNAs related to the epithelial Na/phosphate cotransporter from
zebrafish, mice and flounder.
21,59,60
The transporter, denoted NaPi-II,
is instrumental in maintaining phosphate homeostasis in verte-
brates.
61,62
The antisense transcripts fr
om z
ebrafish have been cloned
and extensively characterized, the mouse homologue(s) has just
recently been detected by PCR and sequenced. The two zebrafish
isoforms (asI and asII) ar
e fully pr
ocessed and encompass most of the
NaPi-II encoding gene. The transcripts lack a significant open reading
frame. RT-PCR experiments revealed a widespread expression of the
antisense transcript in adult zebrafish at a low level. During embryonic
development, however, sense and antisense transcripts are reciprocally
expressed. The antisense transcript is predominantly expressed at all
embryonic stages (3.5 h–2 days), the sense transcript dominates at
day 5 (Werner A, unpublished). The function of the antisense tran-
scripts was investigated by coinjecting both sense and antisense
transcripts in
Xenopus laevis
oocytes. Re-extraction of the injected
RNA followed by northern blotting indicated that both messages
r
emained intact in the cytoplasm of the oocytes.
60
B
idirectional
transcription as such and the NaPi-II encoding transcript are well
conserved between fish and mammals, however, the structure of the
antisense transcripts is highly divergent.
21
The antisense transcript
from mouse which overlaps the phosphate transporter gene shares
the first ex
on with the gene coding for pr
ofilin III located do
wn
-
stream of NaPi-II in opposite orientation.
63
Frq
gene in
Neurospora Crassa.
The non-coding antisense tran-
script of the
frq
gene of
Neurospora crassa
represents one of the few
examples of a naturally occurring antisense RNA with an established
physiological function. The cyclic expression of the
frq
gene is crucial
for rhythmicity in
N
eur
ospora
.
The
fr
q
sense transcript is str
ongly
induced by light and cycles with a period of about 12–14 hours if
grown at constant darkness. The clock can be reset with a light pulse.
An antisense transcript was r
ecently disco
v
er
ed that spans the entir
e
sense transcript. Elegant studies by C. Kramer and S. Crosthwaite
demonstrated the rhythmic expression of both
frq
sense and antisense
transcripts but in opposite phases. Upon knock-out of the antisense
transcript b
y pr
omoter deletion the rhythmic expression of the sense
transcript was still observed, albeit with an delayed periodicity. In
addition, the time course of r
esetting the internal clock after a light
pulse was altered in knock-out strains. It was concluded that the anti-
sense transcript has an impact on setting both the phase and the timing
of the cir
cadian rhythm of
N
eur
ospor
a
in r
esponse external stimuli.
56 RNA Biology 2005; Vol. 2 Issue 2
Natural Antisense Transcripts
Figure 3. (A) Schematic representation of dicer/RISC mediated degradation of perfectly matching, long, double stranded RNAs. In plants and lower eukaryotes
an amplification step by an RNA-dependent RNA polymerase (RdRP) is involved. In mammals, long double-stranded RNAs trigger additional pathways
including an interferon response and RNA editing. (B) Heterochromatic regions with relaxed transcriptional silencing produce overlapping transcripts
containing repetitive sequences. In organisms expressing RdRP the repeats generate binding sites for an efficient amplification of the signal. The duplexes
are processed by dicer and the resulting siRNAs are eventually incorporated into an RNA-induced initiation of transcriptional gene silencing complex (RITS).
RITS is associated with histone methylation and heterochromatin formation. Where the dicing takes place is unclear. It is also unknown to which stage the
duplex is broken down before leaving the nucleus if this was required. (C) MicroRNA synthesis. The primary transcript forms stem loop structures that are
recognized and cleaved by drosha, a nuclear RNase type III. The precursors are transported to the cytoplasm where they are processed further by dicer.
The strands are separated and become integrated into a RISC complex. Its interaction with the target is guided by imperfect base pairing and results in
translational inhibition.
A
B
C
Despite the considerable number of natural antisense transcript
that are functionally characterized a physiological concept is still
impossible to extrapolate. The overlapping genes do not seem to
relate to proteins with specific cellular functions such as oncogenes
or transcription factors, for example.
14,64
The few lessons that can be
extrapolated fr
om the cited examples pr
edict that antisense regulation
r
epr
esents a w
ell
-conserved concept in eukaryotes. In addition, the
often complicated expression pattern of sense and antisense RNA
during development suggests that the consequences of antisense
transcription are more complex than a “simple” downregulation of
the sense transcript.
THE GENOMIC VIEW
The sequencing of the mouse and human genomes combined
with large-scale cDNA sequencing efforts and computational
appr
oaches hav
e r
evealed a clearer picture of extensive bi-directional
transcription. It is now generally accepted that 15 to 25% of
mammalian genes express antisense transcripts.
1,2,14,15,64,65
The
natur
e and the stringency of the parameters that ar
e applied to mine
the various cDNA databanks are responsible for the variation
between predictions.
66
As an example, Yelin et al. excluded non-coding
transcripts that are not spliced and came up with 2667 overlapping
units in the human genome, whereas Chen et al. requested an open
reading frame and/or polyadenylation signal and poly(A) tail. This
approach lead to the pr
ediction of 2940 bi-directionally transcribed
human genes.
2,14
The computational approaches in human comple-
ment efforts pioneered by the Riken group to clone, sequence and
anno
tate full-length cDNAs in mice. They reported 2431 sense/anti-
sense pairs that had a minimal 20 bases overlap in exonic regions
(Fig. 1).
1
The antisense transcripts have been mapped in both mouse
and human
14,15
and were found to be evenly distributed on autosomes.
In human, the numbers varied between 16.4% and 10.1% sense/
antisense pairs on Chrs. 17 and 8, respectively; in mouse the numbers
were 19.7% (Chr. 19) and 12.5% (Chr. 1). Areas where protein
encoding sense transcripts overlap with antisense transcripts were
compared between human and mouse. Orthologous complementary
regions failed to show a significant increase in conservation as com-
pared to non-overlapping areas.
67
This suggests that in mechanistic
terms RNA-RNA interaction and not protein action are relevant for
biological function.
The distribution of overlapping genes on autosomes and the
X
-chromosome, respectively, allows predictions about the mechanism
of gene regulation by natural antisense transcripts. The number of
sense/antisense pairs is significantly lower on the X-chromosome in
both human (7.5%) and mouse (6.3%) as compared to the autosomes.
Antisense transcripts that do not overlap in exonic regions, however,
remain relatively constant on all chromosomes (Fig. 1, bottom
panel).
14
This fact implies that mono-allelic expression of the
X-chromosome selects against an arrangement with overlapping
genes and thus interferes with the regulatory potential of natural
antisense transcripts. This restriction does not apply to overlapping
genes without exonic overlaps. Consequently, a regulatory mechanism
triggered by natural antisense transcripts is likely to involve RNA
hybridization and to act in
trans
.
Two additional phenomena support the assumption that natural
antisense transcripts are transcribed from both alleles: Certain genes
escape X-chromosome inactivation and are expressed from both
alleles. Bi-directionally transcribed genes are clustered in non-silenced
regions of the murine X-chromosome.
68
Furthermore, bi-allelic
expression from the X-chromosome is more frequent in human than
in mouse and so is the occurrence of antisense transcripts.
69
CELLULAR RESPONSES TO DOUBLE STRANDED RNA
I
n eukaryotic cells double stranded RNA triggers a range of
r
eg
ulator
y or pr
otective responses depending on the nature and the
origin of the RNA duplexes. (1) Viral infection potentially leads to
long, perfectly matching double stranded RNA in the cytoplasm of
a cell. The duplex needs to be recognized and eventually destroyed
to guarantee the sur
vival of the cell. RNA interference (RNAi)/post-
transcriptional gene silencing (PTGS) are most prominently
involved at this stage (Fig. 3A). In addition, RNA editing may occur.
70
(2) Repetitive sequences in heterochromatin and transposons show
bi-directional promoter activity that gives rise to double stranded
RNA. Again, a mechanism r
elated to RNAi/PT
GS seems to trigger
protective transcriptional silencing of the relevant DNA region
(Fig. 3B).
71
(3) Single transcripts are able to form intramolecular
stem
str
uctur
es that r
esemble (imper
fectly matched) shor
t double
stranded RNA stretches with a hairpin loop. Some of these structures
are processed by nuclear type III RNases and eventually lead to
regulatory microRNAs (Fig. 3C).
72
(4) Long, imperfectly matched
double stranded RNAs derive from repeat elements and are substrates
for the editing enzyme ADAR (Adenosine deaminase acting on
RNA) (F
ig. 4).
6,73
The four mechanisms all pr
ocess double stranded
RNA, which of the pathways prevails for a given substrate depends
www.landesbioscience.com
RNA
Biology 57
Natural Antisense Transcripts
Figure 4. RNA editing by ADAR. Transcripts that contain low complexity
regions tend to fold back and form long, imperfect stem loop structures. The
duplexes are recognized by ADAR and become partially deaminated. The
edited RNA is retained in the nucleus and eventually degraded. In addition,
the A to I mutations promote unwinding of the duplex.
on the nature of the RNA hybrids and their nuclear/cytoplasmic
location. Redundancy between the pathways seems to occur.
74,75
The four mechanisms are summarized below. However, it is still
speculative, whether they are involved in the processing of natural
antisense RNAs r
esulting from bi-directional transcription.
RNAi and viral defense.
Long, per
fectly matched RNA duplexes
in the cytoplasm r
epresent the likely consequence of a viral infection.
This causes species specific first-line defense mechanisms, of which
RNA interference and the kinase-mediated interferon response are
relevant for this article. Eukaryotic cells are equipped with a cyto-
plasmic RNase III type endonuclease called dicer. The enzyme binds
to double stranded RNA and dices it into small oligonucleotides of
21–23 base pairs (Fig. 3A).
76
These small RNAs can be visualized by
northern blot after challenging cells or cellular extracts with double
stranded RNA. The RNA duplexes are integrated into an RNA-
induced silencing complex (RISC) that targets complementary
RNAs for degradation.
Dicer-mediated degradation of long double stranded RNAs is
observed all eukaryotes. However, in mammalian cells cytoplasmic
RNA duplexes predominantly trigger an immunological response:
77
An RNA-dependent protein kinase (PKR) induces wide spread
transcriptional and translational silencing and an antiviral interferon
response.
78
Double stranded RNA in the nucleus.
Dicer and exogenous
small interfering RNAs are readily detected in the cytoplasm; thus
RNA interference was perceived as a cytoplasmic pathway. Only
recently RNAi related enzymatic activities have been described in
the nucleus. These include siRNA mediated specific degradation of
nuclear RNAs
79
as well as RNA induced transcriptional silencing.
The latter is induced by double stranded RNA in the nucleus and
involves DNA methylation, histone modification and—in repetitive
areas—heterochromatin formation.
9,71,80,81
The relationship is well
established in fission yeast and
Arabidopsis
and recently a similar
mechanism has been found in a specially engineered chicken/
human hybrid cell line
82
and in
Drosophila
.
83
The model predicts
that in the case of reduced silencing spurious transcription of repet-
itive regions occurs. The resulting sense and antisense transcripts
trigger degradation of the duplex in an RNAi-related process that
eventually leads to transcriptional silencing (Fig. 3B).
80
Subsequently, DNA methyltransferases and histone modifying
enzymes are recruited to establish and spread the silenced chromatin
str
uctur
e.
84
microRNAs.
Micro RNAs are well conserved during evolution
and play an important role during development.
85-88
They are
formed from RNA hairpin structures with loops of generally 6–12
bases and a stem of >16 base pairs (Fig. 3C).
72,89
Characteristically,
the double stranded stem is not per
fectly matched.
The loop str
uc
-
tures in the primary transcript are recognized and trimmed by the
nuclear RNAse-III-like enzyme “drosha”.
90
The pre microRNAs are
then integrated into a complex with expor
tin
-
5/ran
-
GTP and
expor
ted to the cytoplasm where the final processing occurs.
91,92
The resulting microRNA-protein complex leads to translational
inhibition of the target mRNA without destroying the template
(F
ig. 3C).
93
The fact that both mice and z
ebrafish dicer knock
-
out
embryos die very early in development is attributed to blocked
microRNA synthesis and corr
oborates the piv
otal role of
microRNAs.
94,95
MicroRNAs have recently emerged as possible downstream
products of natural antisense transcripts. S
ev
eral micr
oRNAs hav
e
been mapped to sense/antisense overlapping regions in mammals
and
Arabidopsis
.
5,96
In addition, RNAi was recently detected in the
nucleus of H
eLa cells.
97
Thus, the machiner
y to pr
ocess perfectly
matched nuclear RNA hybrids seems to be in place. The inherent
redundancy of microRNA and siRNA pathways would allow the
nuclear siRNAs to either feed into a micr
oRNA pathway or trigger
transcriptional silencing.
The regulatory capacity of natural antisense RNAs fits perfectly
into a concept br
ought forward by Mattick et al. It predicts that a
linear relation between DNA and protein, i.e., one gene giving rise
to a single protein (or a limited number of splice forms) would not
pr
oduce enough r
egulator
y output to coor
dinate its functional inte-
gration into a complex organism. This lead to the speculation that
the vast amount of untranslated RNA produced during transcription
could giv
e rise to r
egulator
y output—possibly also in the form of
micr
oRNAs.
98,99
RNA editing.
Primary transcripts that include repetitive elements
may form long imperfect hairpins that are recognized by ADAR
(F
ig. 4).
The members of this enzyme family deaminate adenosine to
inosine, hence its name “adenosine deaminase acting on RNA”.
73
B
ecause inosine pr
efer
entially base-pairs with guanosine the modifi-
cation interferes with the biological properties of the mRNA. Splice
sites could be modified or, if in a coding region, the point mutations
could giv
e rise to alter
ed pr
otein sequences.
70,100-102
I
n addition,
58 RNA Biology 2005; Vol. 2 Issue 2
Natural Antisense Transcripts
Figure 5. Schematic structure of epigentically regulated genes. The first
panel shows the
α−
globin gene cluster with the
LUC7L
gene transcribed in
opposite direction. The indicated deletion removes
HBA1
and
HBQ1
and
relocates
LUC7L
in close proximity to
HBA2
. The region where sense and
antisense potentially overlap is shaded. The intron/exon structure of the
genes is not indicated, the diagram is not in scale. The second panel
represents the imprinted
Igf2r/Air
gene cluster. Imprinted genes are shown
in black, genes with biallelic expression are represented in grey. The anti-
sense transcript air overlaps with the first exon of
Igf2r
and
Mas
, but also
controls the imprinted expression of
Slc22a2
and
Slc22a3
. The third panel
represents the sense and antisense transcripts responsible for X-chromosome
inactivation. The expression of
Xist
results in silencing of the X-chromosome,
Tsix
interferes with
Xist
expression. The promoter region of
Tsix
contains
elements that are required for “counting and choice”, i.e., ensuring that
one chromosome remains active based on random choice between the two
chromosomes.
www.landesbioscience.com
RNA
Biology 59
ADAR-mediated A to I editing destabilizes the RNA-RNA hybrid
and eventually separates the two strands. Most of the 1637 identified
potential targets for RNA editing
103
involve both intronic and exonic
sequences or repetitive elements in untranslated regions. This lead to
the assumption that editing occurs either befor
e or during pre-
mRNA splicing.
104
The RNAi pathway and RNA editing ar
e both triggered by
extended double stranded RNAs.
105
An indication that the two
pathways interfere was recently provided by Tonkin and Bass who
crossed
C. elegans
defective in RNAi with ADAR knock out animals.
74
Interestingly, the behavioural phenotype of the ADAR null mutants
was rescued in animals with an impaired RNAi pathway. This
implies that editing of double stranded RNA prevents degradation
by RNAi.
Overlapping regions of sense and natural antisense transcripts are
potential substrates for ADAR and may be edited.
32,106
ADAR is
predominantly found in the nucleus and acts on RNA before or
during splicing. In contrast, the vast number of antisense transcripts
are fully processed and potentially exported to the cytoplasm. This
makes natural sense/antisense hybrids an unlikely substrate of a
nuclear ADAR response. Nevertheless, a minor fraction of transcripts
encoding the basic fibroblast growth factor (bFGF) in
Xenopus
oocytes were found to be edited.
31,32
The fact that only a small proportion of the FGF-GFG encoding
transcripts are edited
32
implies that modified sequences may be
under represented in cDNA libraries. There is also the possibility
that extensive editing would interfere with the alignment process of
cDNA and genomic sequences. Nevertheless, one would assume that
at least a minor fraction of edited sense/antisense pairs would be
detected in large-scale cDNA sequencing screens. The recent efforts
in bioinformatics, however, have failed to reveal a connection
between antisense transcription and RNA editing.
1,103
Therefore, it
seems unlikely that natural sense/antisense RNA hybrids represent
prominent substrates of the RNA editing machinery.
ANTISENSE RNA IN EPIGENETICS
Prominent examples of naturally occurring antisense transcripts
are related to epigenetic phenomena such as genomic imprinting,
X-chromosome inactivation and the rearrangement of immunoglob-
ulins and T-cell receptors in leukocytes. DNA methylation represents
an additional hallmark of epigenetic silencing and may be a direct
consequence of antisense transcription. F
indings in organisms as
diverse as yeast,
Arabidopsis
and humans indicate that antisense RNA
can potentially induce methylation of the cognate DNA
sequence.
107-109
G
enerally
, the methylation of cytosine r
esidues
influences packaging and accessibility of DNA and, if methylation
affects a promoter region, interferes with the transcription of the
gene.
Introduction of hairpin structures targeting the promoter region
of a stably transfected reporter gene in
Arabidopsis
lead to the methy-
lation of the pr
omoter and the silencing of the r
epor
ter.
Interestingly, the effect was strictly confined to the area covered by
double stranded RNA.
110,111
The presence of 21–24 RNA
oligomers strongly indicates that the RNAi pathway is inv
olv
ed in
guiding the methylation machinery. A similar approach was followed
to demonstrate RNA induced gene silencing in mammalian cells.
Reporter constructs and endogenous genes were targeted with hair-
pin constructs or siRNAs and could be successfully silenced.
112,113
Several CpG islands in the promoter region of the E-cadherin gene
w
ere shown to become methylated in response to siRNA transfection.
The silencing effect of the different oligonucleotides was additive
and resulted in DNA methyltransferase dependant DNA methylation
and decreased mRNA and protein levels. The effect was found to
spread to a neighboring CG rich area (approximately 60 bp) but not
further (150 bp). Similar effects were obtained using hairpin constructs
against erbB2 (transmembrane r
eceptor tyrosine kinase).
113
I
n addi-
tion, the transpor
t of the RNA molecules into the nucleus was
required for gene silencing.
112
A recent report described DNA methylation in mammals
induced by a natural antisense transcript. Tufarelli et al. investigated
a rare form of
α−
thalessaemia brought about by a deletion in the
globin gene locus.
109
The mutation did not affect the
HBA2
gene
but relocated the constitutively active
LUC7L
gene 300 bp down-
stream of
HBA2. LUC7L
gave rise to an antisense transcript that
overlapped both coding and promoter region of
HBA2
(Fig. 6, top
panel). This constellation lead to the methylation of a CpG island in
the promoter and sustained transcriptional silencing of the gene.
The role of the antisense transcript was tested by generating trans-
genic mice that reproduced the mutated genomic arrangement. In a
series of elegant experiments it was found that methylation of the
CpG island was established early in development. The antisense
transcript was shown to act in
cis
and required the transcription of
the sense transcript, consistent with but not necessarily an indication
of a role for dsRNA.
109
Genomic imprinting.
Genomic imprinting describes an epigenetic
modification where only one allele of a specific gene is expressed.
The expression pattern of imprinted genes depends on the parental
origin and has to be reset on the passage through the germ
line.
114,115
In addition, imprinted expression of genes may be tissue
specific. About 70 imprinted genes have been identified so far in
mouse (http://www.mgu.har.mrc.ac.uk/research/imprinting/imprin-
intro.html); a total of about 200 mammalian imprinted genes are
expected. Imprinted genes are arranged in clusters and the promoter
regions of maternal and paternal alleles are often differentially
methylated. So-called imprinting control centers (IC) keep the
imprinted clusters under coordinate control. Of interest for this
report is the fact that every imprinted gene cluster seems to have an
antisense transcript that plays a piv
otal r
ole in imprinting.
11,116
The best characterized imprinted sense/ antisense pair encode the
insulin like growth factor 2 receptor (
Igf2r
) and the cognate anti-
sense transcript
Air
(Fig. 5).
11,117,118
The expression of the antisense
RNA is controlled by a differentially methylated region located in
the second intr
on of the sense transcript.
Air
is paternally transcribed
and overlaps the 5

end of the
Igf2r
gene. Expression of
Air
induces
the mono-allelic maternal expression of the Igf2 receptor as well as
S
lc22a2/S
lc22a3
that do not o
v
erlap with Air
. A transgenic mouse
that lacks the differentially methylated control center shows bi-allelic
expr
ession of the
I
g
f2r/S
lc22a2/Slc22a3
cluster
.
25
T
o r
ule out a dir
ect
influence of the deletion transcription of
Air
was prematurely termi-
nated by the insertion of a polyadenylation cassette. The mice
sho
w
ed lo
w
er bir
th w
eight characteristic for bi
-allelic expression of
Igf2r
. However, when the 5’ end of the
Igf2r
gene was modified to
remove any overlap with
Air
imprinting was not impaired.
25
These
experiments showed that the antisense transcript was essential in
maintaining the imprinted cluster but questioned the r
ole of the
RNA-RNA overlap.
119
X
-
chromosome inactivation.
X
-
chromosome inactivation describes
a mechanism in mammals that aims to silence X-chromosomes with
the purpose of dosage compensation.
The process is initiated and
Natural Antisense Transcripts

60 RNA Biology 2005; Vol. 2 Issue 2
controlled at the X-inactivation center and involves various species
specific regulatory elements including the overlapping non-coding
RNA
Xist
(X-inactive specific transcript) and the cognate antisense
RNA
Tsix
(Fig. 5, bottom panel).
120-122
Xist
recruits the components
of the pr
otein machinery that induces CpG methylation, histone
modification and ev
entually leads to heterochromatization of the
X
-chromosome.
123
The expr
ession of
Xist
is r
egulated by
T
six
, which
overlaps
Xist
in antisense direction.
124-126
Premature transcriptional
termination of
Tsix
leads to the obligatory expression of
Xist
and
concomitant silencing of the chromosome that lacks
Tsix
. A
Tsix
transgene in reverse orientation (resulting in a head to tail arrange-
ment of
Xist
and
Tsix
as compared to the naturally occurring gene
overlap) was unable to restore random X-chromosome inactivation.
13
This indicated that transcription across the
Xist
locus was required
for correct silencing. Accordingly,
Tsix
is only active in a
cis
-acting
mechanism.
Genomic recombination.
Extensive antisense transcription was
observed during VDJ
H
recombination of the immunoglobulin heavy
chain locus. The expression of the antisense transcripts was strictly
confined to the V-region, involved genic and intergenic regions and
stopped immediately after VDJ
H
recombination.
127
Transcription
was observed at both alleles and it was hypothesized that the antisense
RNA may act as a chromatin opener or a scaffold for the recombi-
nation machinery. After the recombination event one of the alleles is
silenced to guarantee the clonal character of mature B-and T-cells.
The silencing step is likely to happen at the trailing allele and to
include additional protein components.
To summarize, antisense transcription in the latter three cases
appears to be closely linked to mono-allelic expression as observed in
imprinting, X-chromosome inactivation and genomic rearrangements
in leukocytes. All these phenomena have evolved relatively late in
evolution; in contrast, regulatory antisense RNAs seem to represent
a rather ancient concept.
CONCLUSION
Antisense transcription represents a widespread phenomenon in
eukaryotes. The cellular consequences of the resulting potentially
hybridizing transcripts are rather speculative and the mechanistic
concept is based on indirect evidence. The genomic research indicates
that a large proportion of natural antisense transcripts require
bi-allelic expression to exhibit biological function. In addition, an
o
v
erlap of sense and antisense transcript seems to be necessary.
Activation of an RNAi related mechanism is a likely consequence
possibly leading to the formation siRNA/microRNAs with signaling
function.
The antisense transcripts found in gene clusters that are
mono
-
allelically expr
essed ar
e suggested have a different function.
Their mode of action does not require an overlap with the sense
transcript. Antisense transcription triggers the silencing of wider
genomic ar
eas in
cis
.
The antisense RNA is suggested to function as
a scaffold and r
ecruit the protein machinery to mediate transcriptional
silencing.
References
1.Okazaki Y, Furuno M, Kasukawa T, Adachi J, Bono H, Kondo S, Nikaido I, Osato N, Saito
R, Suzuki H, Yamanaka I, Kiyosawa H, Yagi K, Tomaru Y, Hasegawa Y, Nogami A,
Schonbach C, G
ojobori T, Baldarelli R, Hill DP, Bult C, Hume DA, Quackenbush J,
Schriml LM, Kanapin A, M
atsuda H, Batalov S, Beisel KW, Blake JA, Bradt D, Brusic V,
Chothia C, Corbani LE, Cousins S, D
alla E, Dragani TA, Fletcher CF, Forrest A, Frazer
KS, Gaasterland T, Gariboldi M, Gissi C, Godzik A, Gough J, Grimmond S, Gustincich
S, Hirokawa N, Jackson IJ, Jarvis ED, Kanai A, Kawaji H, Kawasawa Y, Kedzierski RM,
King BL, K
onagaya A, Kurochkin IV, Lee Y, Lenhard B, Lyons PA, Maglott DR, Maltais
L, M
archionni L, McKenzie L, Miki H, Nagashima T, Numata K, Okido T, Pavan WJ,
P
ertea G, Pesole G, Petrovsky N, Pillai R, Pontius JU, Qi D, Ramachandran S, Ravasi T,
Reed JC, Reed DJ, Reid J, Ring BZ, Ringwald M, Sandelin A, Schneider C, Semple CA,
Setou M, Shimada K, Sultana R, Takenaka Y, Taylor MS, Teasdale RD, Tomita M, Verardo
R,
Wagner L, Wahlestedt C, Wang Y, Watanabe Y, Wells C, Wilming LG, Wynshaw-Boris
A,
Yanagisawa M, Yang I, Yang L, Yuan Z, Zavolan M, Zhu Y, Zimmer A, Carninci P,
H
ayatsu N, Hirozane-Kishikawa T, Konno H, Nakamura M, Sakazume N, Sato K, Shiraki
T, Waki K, Kawai J, Aizawa K, Arakawa T, Fukuda S, Hara A, Hashizume W, Imotani K,
Ishii Y, Itoh M, Kagawa I, Miyazaki A, Sakai K, Sasaki D, Shibata K, Shinagawa A,
Y
asunishi A, Yoshino M, Waterston R, Lander ES, Rogers J, Birney E, Hayashizaki Y; FAN-
T
OM Consortium; RIKEN Genome Exploration Research Group Phase I & II Team.
Analysis of the mouse transcriptome based on functional annotation of 60,770 full
-length
cDNAs. Nature 2002; 420:563-73.
2.Yelin R, Dahary D, Sorek R, Levanon EY, Goldstein O, Shoshan A, Diber A, Biton S,
Tamir Y, Khosravi R, Nemzer S, Pinner E, Walach S, Bernstein J, Savitsky K, Rotman G.
W
idespread occurrence of antisense transcription in the human genome. Nat Biotechnol
2003; 21:379-86.
3.
Misra S, Crosby MA, Mungall CJ, Matthews BB, Campbell KS, Hradecky P, Huang Y,
Kaminker JS, Millburn GH, Prochnik SE, Smith CD, Tupy JL, Whitfied EJ, Bayraktaroglu
L, Berman BP, Bettencour
t BR, Celniker SE, de Grey AD, Drysdale RA, Harris NL,
Richter J, R
usso S, Schroeder AJ, Shu SQ, Stapleton M, Yamada C, Ashburner M, Gelbart
WM, Rubin GM, Lewis SE. Annotation of the Drosophila melanogaster euchromatic
genome: A systematic r
eview. Genome Biol 2002; 3:RESEARCH0083.
4.Osato N, Yamada H, Satoh K, Ooka H, Yamamoto M, Suzuki K, Kawai J, Carninci P,
Ohtomo Y, Murakami K, Matsubara K, Kikuchi S, Hayashizaki Y. Antisense transcripts
with rice full-length cDNAs. Genome Biol 2003; 5:R5.
5.Wang XJ, Gaasterland T, Chua NH. Genome-wide prediction and identification of cis-nat-
ural antisense transcripts in Arabidopsis thaliana. Genome Biol 2005; 6:R30.
6.Wang Q, Carmichael GG. Effects of length and location on the cellular response to dou-
ble-stranded RNA. Microbiol Mol Biol Rev 2004; 68:432-52, (table of contents).
7.Mello CC, Conte Jr D. Revealing the world of RNA interference. Nature 2004; 431:338-42.
8.O’Neill MJ. The influence of noncoding RNAs on allele-specific gene expression in mam-
mals. Hum Mol Genet 2005; 14:R113-20.
9.Matzke M, Aufsatz W, Kanno T, Daxinger L, Papp I, Mette MF, Matzke AJ. Genetic analy-
sis of RNA-mediated transcriptional gene silencing. Biochim Biophys Acta 2004;
1677:129-41.
10.Malik K, Yan P, Huang TH, Brown KW. Wilms’ tumor: A paradigm for the new genetics.
Oncol Res 2000; 12:441-9.
11.Rougeulle C, Heard E. Antisense RNA in imprinting: Spreading silence through Air.
Trends Genet 2002; 18:434-7.
12.Ogawa Y, Lee JT. Antisense regulation in X inactivation and autosomal imprinting.
Cytogenet Genome Res 2002; 99:59-65.
13.Shibata S, Lee JT. Tsix transcription- versus RNA-based mechanisms in Xist repression and
epigenetic choice. C
urr Biol 2004; 14:1747-54.
14.Chen J, Sun M, Kent WJ, Huang X, Xie H, Wang W, Zhou G, Shi RZ, Rowley JD. Over
20% of human transcripts might form sense
-antisense pairs. Nucleic Acids Res 2004;
32:4812-20.
15.
Kiyosawa H, Yamanaka I, Osato N, Kondo S, Hayashizaki Y. Antisense transcripts with
F
ANT
OM2 clone set and their implications for gene regulation. Genome Res 2003;
13:1324-34.
16.
Knee R, M
urphy PR. Regulation of gene expression by natural antisense RNA transcripts.
Neurochem Int 1997; 31:379-92.
17.
D
olnick BJ. Naturally occurring antisense RNA. Pharmacol Ther 1997; 75:179-84.
18.Kumar M, Carmichael GG. Antisense RNA: Function and fate of duplex RNA in cells of
higher eukar
yotes. Microbiol Mol Biol Rev 1998; 62:1415-34.
19.Lipman DJ. Making (anti)sense of noncoding sequence conservation. Nucleic Acids Res
1997; 25:3580-3.
20.Vanhee-Brossollet C, Vaquero C. Do natural antisense transcripts make sense in eukary-
otes? G
ene 1998; 211:1-9.
21.Werner A, Preston-Fayers K, Dehmelt L, Nalbant P. Regulation of the NPT gene by a nat-
urally occurring antisense transcript. Cell B
iochem Biophys 2002; 36:241-52.
22.Nellen W, Lichtenstein C. What makes an mRNA anti-sense-itive? Trends Biochem Sci
1993; 18:419-23.
23.Wagner EG, Flardh K. Antisense RNAs everywhere? Trends Genet 2002; 18:223-6.
24.
Cawley S, B
ekirano
v S, Ng HH, Kaprano
v P
, S
ekinger EA, Kampa D, P
iccolboni A,
Sementchenko V, Cheng J, Williams AJ, Wheeler R, Wong B, Drenkow J, Yamanaka M,
Patel S, Brubaker S, Tammana H, Helt G, Struhl K, Gingeras TR. Unbiased mapping of
transcription factor binding sites along human chr
omosomes 21 and 22 points to wide
-
spr
ead r
egulation of noncoding RNAs. Cell 2004; 116:499-509.
Natural Antisense Transcripts

www.landesbioscience.com
RNA
Biology 61
25.
Sleutels F, Zwart R, Barlow DP. The noncoding Air RNA is required for silencing autoso-
mal imprinted genes. N
ature 2002; 415:810-813.
26.
Lee JT, Lu N. Targeted mutagenesis of Tsix leads to nonrandom X inactivation. Cell 1999;
99:47-57.
27.
Kramer C, Loros JJ, Dunlap JC, Crosthwaite SK. Role for antisense RNA in regulating cir-
cadian clock function in N
eurospora crassa. Nature 2003; 421:948-52.
28.
Crosthwaite SK. Circadian clocks and natural antisense RNA. FEBS Lett 2004; 567:49-54.
29.Gagnon ML, Moy GK, Klagsbrun M. Characterization of the promoter for the human
antisense fibroblast growth factor-2 gene; Regulation by Ets in Jurkat T cells. J Cell
B
iochem 1999; 72:492-506.
30.
Mihalich A, Reina M, Mangioni S, Ponti E, Alberti L, Vigano P, Vignali M, Di Blasio AM.
Different basic fibroblast growth factor and fibroblast growth factor-antisense expression in
eutopic endometrial stromal cells derived from women with and without endometriosis. J
Clin E
ndocrinol Metab 2003; 88:2853-9.
31.
Kimelman D, Kirschner MW. An antisense mRNA directs the covalent modification of the
transcript encoding fibroblast growth factor in Xenopus oocytes. Cell 1989; 59:687-96.
32.Saccomanno L, Bass BL. A minor fraction of basic fibroblast growth factor mRNA is deam-
inated in X
enopus stage VI and matured oocytes. Rna 1999; 5:39-48.
33.
Savage MP, Fallon JF. FGF-2 mRNA and its antisense message are expressed in a develop-
mentally specific manner in the chick limb bud and mesonephros. Dev Dyn 1995;
202:343-53.
34.Murphy PR, Knee RS. Identification and characterization of an antisense RNA transcript
(g
fg) from the human basic fibroblast growth factor gene. Mol Endocrinol 1994; 8:852-9.
35.
Li A
W, Too CK, Murphy PR. The basic fibroblast growth factor (FGF-2) antisense RNA
(GFG) is translated into a MutT-related protein in vivo. Biochem Biophys Res Commun
1996; 223:19-23.
36.
Asa SL, Ramyar L, Murphy PR, Li AW, Ezzat S. The endogenous fibroblast growth factor-2
antisense gene product regulates pituitary cell growth and hormone production. Mol
E
ndocrinol 2001; 15:589-99.
37.Li AW, Murphy PR. Expression of alternatively spliced FGF-2 antisense RNA transcripts
in the central nervous system: Regulation of FGF-2 mRNA translation. Mol Cell
Endocrinol 2000; 170:233-42.
38.Grothe C, Meisinger C. Fibroblast growth factor (FGF)-2 sense and antisense mRNA and
FGF receptor type 1 mRNA are present in the embryonic and adult rat nervous system:
Specific detection by nuclease protection assay. Neurosci Lett 1995; 197:175-8.
39.Rossignol F, Vache C, Clottes E. Natural antisense transcripts of hypoxia-inducible factor
1alpha are detected in different normal and tumour human tissues. Gene 2002; 299:135-40.
40.Rossignol F, De Laplanche E, Mounier R, Bonnefont J, Cayre A, Godinot C, Simonnet H,
Clottes E. Natural antisense transcripts of HIF-1alpha are conserved in rodents. Gene
2004; 339:121-30.
41.Uchida T, Rossignol F, Matthay MA, Mounier R, Couette S, Clottes E, Clerici C.
Prolonged hypoxia differentially regulates hypoxia-inducible factor (HIF)-1alpha and
HIF-2alpha expression in lung epithelial cells: Implication of natural antisense HIF-1alpha.
J Biol Chem 2004; 279:14871-8.
42.Cayre A, Rossignol F, Clottes E, Penault-Llorca F. aHIF but not HIF-1alpha transcript is a
poor prognostic marker in human breast cancer. Breast Cancer Res 2003; 5:R223-30.
43.Thrash-Bingham CA, Tartof KD. aHIF: A natural antisense transcript overexpressed in
human renal cancer and during hypoxia. J Natl Cancer Inst 1999; 91:143-51.
44.Sharpe PT. Homeobox genes and orofacial development. Connect Tissue Res 1995; 32:17-25.
45.
B
lin-Wakkach C, Lezot F, Ghoul-Mazgar S, Hotton D, Monteiro S, Teillaud C, Pibouin L,
O
r
estes-Car
doso S, Papagerakis P, Macdougall M, Robert B, Berdal A. Endogenous Msx1
antisense transcript: In vivo and in vitro evidences, structure, and potential involvement in
skeleton dev
elopment in mammals. P
roc Natl Acad Sci USA 2001; 98:7336-41.
46.
Le
z
ot F
, Couder
t A, P
etit S,
Vi-Fane B, Hotton D, Davideau Jl, Kato S, Descroix V, Pibouin
L, B
er
dal A. Does Vitamin D play a role on Msx1 homeoprotein expression involving an
endogenous antisense mRNA? J S
ter
oid B
iochem Mol Biol 2004; 89-90:413-7.
47.Zhang J, Lazar MA. The mechanism of action of thyroid hormones. Annu Rev Physiol
2000; 62:439-66.
48.Lazar MA. Thyroid hormone receptors: Multiple forms, multiple possibilities. Endocr Rev
1993; 14:184-93.
49.Laudet V, Begue A, Henry-Duthoit C, Joubel A, Martin P, Stehelin D, Saule S. Genomic
organization of the human thyroid hormone receptor alpha (c-erbA-1) gene. Nucleic Acids
Res 1991; 19:1105-12.
50.Miyajima N, Horiuchi R, Shibuya Y, Fukushige S, Matsubara K, Toyoshima K, Yamamoto
T
.
T
wo erbA homologs encoding pr
oteins with different T3 binding capacities are tran-
scribed from opposite DNA strands of the same genetic locus. Cell 1989; 57:31-9.
51.
Lazar MA, H
odin RA, D
arling DS, Chin
WW. A novel member of the thyroid/steroid hor-
mone receptor family is encoded by the opposite strand of the rat c-erbA alpha transcrip-
tional unit. M
ol Cell B
iol 1989; 9:112836.
52.Hastings ML, Ingle HA, Lazar MA, Munroe SH. Post-transcriptional regulation of thyroid
hormone receptor expression b
y cis
-
acting sequences and a naturally occurring antisense
RNA. J Biol Chem 2000; 275:11507-13.
53.Hastings ML, Milcarek C, Martincic K, Peterson ML, Munroe SH. Expression of the thy-
roid hormone receptor gene, erbAalpha, in B lymphocytes: Alternative mRNA processing
is independent of differentiation but correlates with antisense RNA levels. Nucleic Acids
Res 1997; 25:4296-300.
54.Munroe SH, Lazar MA. Inhibition of c-erbA mRNA splicing by a naturally occurring anti-
sense RNA. J Biol Chem 1991; 266:22083-6.
55.
Luther HP, Podlowski S, Hetzer R, Baumann G. Analysis of sense and naturally occurring
antisense transcripts of my
osin heavy chain in the human myocardium. J Cell Biochem
2001; 80:596-605.
56.
Luther HP, Haase H, Hohaus A, Beckmann G, Reich J, Morano I. Characterization of nat-
urally occurring myosin heavy chain antisense mRNA in rat heart. J Cell Biochem 1998;
70:110-20.
57.
Haddad F, Bodell PW, Qin AX, Giger JM, Baldwin KM. Role of antisense RNA in coor-
dinating car
diac myosin heavy chain gene switching. J Biol Chem 2003; 278:37132-8.
58.Rindt H, Gulick J, Knotts S, Neumann J, Robbins J. In vivo analysis of the murine
beta-myosin heavy chain gene promoter. J Biol Chem 1993; 268:5332-8.
59.
Nalbant P, Boehmer C, Dehmelt L, Wehner F, Werner A. Functional characterization of a
N
a+-phosphate cotransporter (NaPi-II) from zebrafish and identification of related tran-
scripts. J P
hysiol 1999; 520:79-89.
60.Huelseweh B, Kohl B, Hentschel H, Kinne RK, Werner A. Translated anti-sense product
of the Na/phosphate cotransporter (NaPi-II). Biochem J 1998; 332:483-9.
61.
Murer H, Hernando N, Forster I, Biber J. Regulation of Na/Pi transporter in the proximal
tubule. Annu R
ev Physiol 2003; 65:531-42.
62.Werner A, Kinne RK. Evolution of the Na-P(i) cotransport systems. Am J Physiol Regul
Integr Comp Physiol 2001; 280:R301-12.
63.Braun A, Aszodi A, Hellebrand H, Berna A, Fassler R, Brandau O. Genomic organization
of pr
ofilin-III and evidence for a transcript expressed exclusively in testis. Gene 2002;
283:219-25.
64.Shendure J, Church GM. Computational discovery of sense-antisense transcription in the
human and mouse genomes. Genome Biol 2002; 3:RESEARCH0044.
65.Lehner B, Williams G, Campbell RD, Sanderson CM. Antisense transcripts in the human
genome.
Trends Genet 2002; 18:63-5.
66.
Lavorgna G, Dahary D, Lehner B, Sorek R, Sanderson CM, Casari G. In search of anti-
sense. Trends in Biochemical Sciences 2004; 29:88-94.
67.Veeramachaneni V, Makalowski W, Galdzicki M, Sood R, Makalowska I. Mammalian over-
lapping genes: The comparative perspective. Genome Res 2004; 14:280-6.
68.Kiyosawa H, Abe K. Speculations on the role of natural antisense transcripts in mammalian
X chromosome evolution. Cytogenet Genome Res 2002; 99:151-6.
69.Brown CJ, Greally JM. A stain upon the silence: Genes escaping X inactivation. Trends
Genet 2003; 19:432-8.
70.Saunders LR, Barber GN. The dsRNA binding protein family: Critical roles, diverse cellu-
lar functions. Faseb J 2003; 17:961-83.
71.Grewal SI, Rice JC. Regulation of heterochromatin by histone methylation and small
RNAs. Curr Opin Cell Biol 2004; 16:230-8.
72.Bartel DP. MicroRNAs: Genomics, biogenesis, mechanism, and function. Cell 2004;
116:281-97.
73.Maas S, Rich A, Nishikura K. A-to-I RNA editing: Recent news and residual mysteries. J
Biol Chem 2003; 278:1391-4.
74.Tonkin LA, Bass BL. Mutations in RNAi rescue aberrant chemotaxis of ADAR mutants.
Science 2003; 302:1725.
75.Tang G. siRNA and miRNA: An insight into RISCs. Trends Biochem Sci 2005; 30:106-14.
76.Zamore PD, Tuschl T, Sharp PA, Bartel DP. RNAi: Double-stranded RNA directs the
ATP-dependent cleavage of mRNA at 21 to 23 nucleotide intervals. Cell 2000; 101:25-33.
77.Yang S, Tutton S, Pierce E, Yoon K. Specific double-stranded RNA interference in undif-
ferentiated mouse embryonic stem cells. Mol Cell Biol 2001; 21:7807-16.
78.
Clemens MJ. PKR—a pr
otein kinase r
egulated by double-stranded RNA. Int J Biochem
Cell Biol 1997; 29:945-9.
79.Robb GB, Brown KM, Khurana J, Rana TM. Specific and potent RNAi in the nucleus of
human cells. N
at S
truct Mol Biol 2005; 12:133-7.
80.Martienssen RA. Maintenance of heterochromatin by RNA interference of tandem repeats.
Nat Genet 2003; 35:213-4.
81.
Lippman Z, Martienssen R. The role of RNA interference in heterochromatic silencing.
Nature 2004; 431:364-70.
82.Fukagawa T, Nogami M, Yoshikawa M, Ikeno M, Okazaki T, Takami Y, Nakayama T,
Oshimura M. Dicer is essential for formation of the heterochromatin structure in verte-
brate cells. Nat Cell Biol 2004; 6:784-91.
83.
P
al
-
B
hadra M, Leibo
vitch BA, G
andhi SG, Rao M, Bhadra U, Birchler JA, Elgin SC.
Heterochromatic silencing and HP1 localization in Drosophila are dependent on the RNAi
machinery. Science 2004; 303:669-72.
84.Verdel A, Jia S, Gerber S, Sugiyama T, Gygi S, Grewal SI, Moazed D. RNAi-mediated tar-
geting of heter
ochr
omatin b
y the RIT
S complex. Science 2004; 303:672-6.
85.Pasquinelli AE, Mccoy A, Jimenez E, Salo E, Ruvkun G, Martindale MQ, Baguna J.
Expression of the 22 nucleotide let-7 heterochronic RNA throughout the Metazoa: A role
in life history evolution? Evol Dev 2003; 5:372-8.
86.
P
asquinelli AE, Reinhart BJ, Slack F, Martindale MQ, Kuroda MI, Maller B, Hayward DC,
Ball EE, Degnan B, Muller P, Spring J, Srinivasan A, Fishman M, Finnerty J, Corbo J,
Levine M, Leahy P, Davidson E, Ruvkun G. Conservation of the sequence and temporal
expression of let-7 heterochronic regulatory RNA. Nature 2000; 408:86-9.
87.
R
einhart BJ, Slack FJ, Basson M, Pasquinelli AE, Bettinger JC, Rougvie AE, Horvitz HR,
Ruvkun G. The 21-nucleotide let-7 RNA regulates developmental timing in
Caenorhabditis elegans. Nature 2000; 403:901-6.
88.
M
oss EG,
Tang L. Conservation of the heterochronic regulator Lin-28, its developmental
expr
ession and microRNA complementary sites. Dev Biol 2003; 258:432-42.
89.Ambros V. The functions of animal microRNAs. Nature 2004; 431:350-5.
Natural Antisense Transcripts

62 RNA Biology 2005; Vol. 2 Issue 2
90.
Lee Y, Ahn C, Han J, Choi H, Kim J, Yim J, Lee J, Provost P, Radmark O, Kim S, Kim
V
N. The nuclear RNase III Drosha initiates microRNA processing. Nature 2003; 425:415-9.
91.
Yi R, Qin Y, Macara IG, Cullen BR. Exportin-5 mediates the nuclear export of
premicroRNAs and short hairpin RNAs. Genes Dev 2003; 17:3011-6.
92.Zeng Y, Cullen BR. Structural requirements for premicroRNA binding and nuclear export
b
y Exportin 5. Nucleic Acids Res 2004; 32:4776-85.
93.
Olsen PH, Ambros V. The lin-4 regulatory RNA controls developmental timing in
Caenorhabditis elegans b
y blocking LIN-14 protein synthesis after the initiation of trans-
lation. D
ev Biol 1999; 216:671-80.
94.
Bernstein E, Kim SY, Carmell MA, Murchison EP, Alcorn H, LI MZ, Mills AA, Elledge
SJ, Anderson KV, Hannon GJ. Dicer is essential for mouse development. Nat Genet 2003;
35:215-7.
95.
Wienholds E, Koudijs MJ, Van Eeden FJ, Cuppen E, Plasterk RH. The microRNA-pro-
ducing enzyme D
icer1 is essential for zebrafish development. Nat Genet 2003; 35:217-8.
96.
Davis E, Caiment F, Tordoir X, Cavaille J, Ferguson-Smith A, Cockett N, Georges M,
Charlier C. RNAi
-mediated allelic trans-interaction at the imprinted Rtl1/Peg11 locus.
Curr Biol 2005; 15:743-9.
97.Robb GB, Carson AR, Tai SC, Fish JE, Singh S, Yamada T, Scherer SW, Nakabayashi K,
M
arsden PA. Post-transcriptional regulation of endothelial nitric-oxide synthase by an
o
verlapping antisense mRNA transcript. J Biol Chem 2004; 279:37982-96.
98.
Mattick JS. RNA regulation: A new genetics? Nat Rev Genet 2004; 5:316-23.
99.
Herbert A. The four Rs of RNA-directed evolution. Nat Genet 2004; 36:19-25.
100.Bass BL. RNA editing and hypermutation by adenosine deamination. Trends Biochem Sci
1997; 22:157-62.
101.Bass BL. RNA editing by adenosine deaminases that act on RNA. Annu Rev Biochem
2002; 71:817-46.
102.
Seeburg PH, Hartner J. Regulation of ion channel/neurotransmitter receptor function by
RNA editing. C
urr Opin Neurobiol 2003; 13:279-83.
103.Levanon EY, Eisenberg E, Yelin R, Nemzer S, Hallegger M, Shemesh R, Fligelman ZY,
Shoshan A, Pollock SR, Sztybel D, Olshansky M, Rechavi G, Jantsch MF. Systematic iden-
tification of abundant A
-to-I editing sites in the human transcriptome. Nat Biotechnol
2004; 22:1001-5.
104.Raitskin O, Cho DS, Sperling J, Nishikura K, Sperling R. RNA editing activity is associ-
ated with splicing factors in lnRNP particles: The nuclear premRNA processing machin-
ery. Proc Natl Acad Sci USA 2001; 98:6571-6.
105.Knight SW, Bass BL. The role of RNA editing by ADARs in RNAi. Mol Cell 2002;
10:809-17.
106.Zhang Z, Carmichael GG. The fate of dsRNA in the nucleus: A p54(nrb)-containing com-
plex mediates the nuclear retention of promiscuously A-to-I edited RNAs. Cell 2001;
106:465-75.
107.Grewal SI, Moazed D. Heterochromatin and epigenetic control of gene expression. Science
2003; 301:798-802.
108.Matzke M, Matzke AJ, Kooter JM. RNA: Guiding gene silencing. Science 2001;
293:1080-3.
109.Tufarelli C, Stanley JA, Garrick D, Sharpe JA, Ayyub H, Wood WG, Higgs DR.
Transcription of antisense RNA leading to gene silencing and methylation as a novel cause
of human genetic disease. Nat Genet 2003; 34:157-65.
110.Mette MF, Aufsatz W, Van Der Winden J, Matzke MA, Matzke AJ. Transcriptional silencing
and promoter methylation trigger
ed b
y double-stranded RNA. Embo J 2000; 19:5194-201.
111.Aufsatz W, Mette MF, Van Der Winden J, Matzke AJ, Matzke M. RNA-directed DNA
methylation in Arabidopsis. Proc Natl Acad Sci USA 2002; 99:16499-506.
112.Morris KV, Chan SW, Jacobsen SE, Looney DJ. Small interfering RNA-induced transcrip-
tional gene silencing in human cells. Science 2004; 305:1289-92.
113.Kawasaki H, Taira K. Induction of DNA methylation and gene silencing by short interfer-
ing RNAs in human cells. Nature 2004; 431:211-7.
114.Reik W, Dean W, Walter J. Epigenetic reprogramming in mammalian development.
Science 2001; 293:1089-93.
115.Jaenisch R, Bird A. Epigenetic regulation of gene expression: How the genome integrates
intrinsic and environmental signals. Nat Genet 2003; 33:245-54.
116.Malik K, Brown KW. Epigenetic gene deregulation in cancer. Br J Cancer 2000; 83:1583-8.
117.
S
leutels F
, Barlow DP, Lyle R. The uniqueness of the imprinting mechanism. Curr Opin
Genet Dev 2000; 10:229-33.
118.
W
utz A, S
mrzka OW, Barlow DP. Making sense of imprinting the mouse and human
IGF2R loci. N
o
v
ar
tis Found Symp 1998; 214:251-9, (discussion 260-53).
119.
S
leutels F, Tjon G, Ludwig T, Barlow DP. Imprinted silencing of Slc22a2 and Slc22a3 does
not need transcriptional o
v
erlap betw
een I
gf2r and Air. Embo J 2003; 22:3696-704.
120.
O
gawa Y, Lee JT. Xite, X-inactivation intergenic transcription elements that regulate the
pr
obability of choice. M
ol Cell 2003; 11:731-43.
121.
B
rown CJ, Chow JC. Beyond sense: The role of antisense RNA in controlling Xist expres-
sion. Semin Cell Dev Biol 2003; 14:341-7.
122.Heard E. Recent advances in X-chromosome inactivation. Curr Opin Cell Biol 2004;
16:247-55.
123.Brockdorff N. X-chromosome inactivation: Closing in on proteins that bind Xist RNA.
Trends Genet 2002; 18:352-8.
124.Shibata S, Lee JT. Characterization and quantitation of differential Tsix transcripts:
Implications for Tsix function. Hum Mol Genet 2003; 12:125-36.
125.
Lee JT, Davidow LS, Warshawsky D. Tsix, a gene antisense to Xist at the X-inactivation
centre. Nat Genet 1999; 21:400-4.
126.Nesterova TB, Johnston CM, Appanah R, Newall AE, Godwin J, Alexiou M, Brockdorff
N. S
kewing X chromosome choice by modulating sense transcription across the Xist locus.
G
enes Dev 2003; 17:2177-90.
127.
Bolland DJ, Wood AL, Johnston CM, Bunting SF, Morgan G, Chakalova L, Fraser PJ,
Cor
coran AE. Antisense intergenic transcription in V(D)J recombination. Nat Immunol
2004; 5:630-7.
Natural Antisense Transcripts