mRNA Editing

mRNA Editing Advanced article

Christopher L Sansam, Vanderbilt University, Nashville, Tennessee, USA Article contents

Ronald B Emeson, Vanderbilt University, Nashville, Tennessee, USA
Introduction

C-to-U Editing
The editing of messenger RNA transcripts respresents a recently identified processing
A-to-I Editing
event by which multiple RNA transcripts can be generated from a single genomic locus to
Perspectives and Future Directions
increase the coding potential of the human genome.
doi: 10.1038/npg.els.0005041

Introduction
C-to-U Editing
Although completion of the Human Genome Project
represents a milestone in modern biology, it has
become increasingly apparent that the primary nucle-
Apolipoprotein B
otide sequence of human genetic material cannot be The first editing event identified in human RNA was a
used to predict accurately the nucleotide sequences of single C-to-U conversion in mRNA transcripts en-
many expressed ribonucleic acids (RNAs) or the coding apolipoprotein B (apoB; Powell et al., 1987).
amino acid sequences of their encoded protein Two distinct proteins, apoB-100 and apoB-48, can be
products. RNAs encoded by most eukaryotic genes produced from the apolipoprotein B (APOB) gene
can undergo different posttranscriptional RNA pro- through a tissue-specific C-to-U editing event at
cessing events (splicing, capping, polyadenylation), cytidine 6666 in APOB transcripts, which converts a
which are required to convert RNA precursors into genomically encoded glutamine codon (CAA) to an
mature RNA transcripts. During the mid-1980s, an in-frame stop codon (UAA; Figure 2a). This single-
additional RNA processing term, called ‘RNA base modification allows the production of apoB-48 in
editing’, was coined to describe a genetic phenomenon the small intestine; this apoB variant of 241 kilo-
in which uridine residues were inserted and deleted daltons (kDa) is a component of the triglyceride-rich
from the mitochondrial RNAs of kinetoplastid
protozoa (Benne et al., 1986). Since then, the term
RNA editing has been used to describe numerous NH2 O
cellular processes resulting in the modification of RNA H2O NH3
sequences differing from that designated by their
deoxyribonucleic acid (DNA) or RNA templates in N NH
organisms as diverse as bacteria, protozoa, plants and
humans. These processes, which affect messenger N O N O
RNAs (mRNAs), transfer RNAs (tRNAs) and ribo-
somal RNAs, can alter the structure, stability, R R
translation efficiency, splicing patterns and coding Cytidine Uridine
potential of the modified transcripts, thereby affecting
almost all aspects of cellular RNA function. (See RNA
NH2 O
Processing; Splicing of pre-mRNA.) H2O NH3
So far, the identification of editing events in human N N
RNAs has been limited largely to nucleoside modifi- N NH
cations in which specific cytidine residues are
posttranscriptionally converted to uridine moieties N N
N
N
(C-to-U editing) or in which specific adenosines are
R R
converted to inosines (A-to-I editing; Figure 1). C-to-
U and A-to-I editing events are mediated by distinct Adenosine Inosine
mechanisms that require specific regulatory sequence
Figure 1 RNA editing by hydrolytic deamination. RNA editing
elements and cellular factors, but both types of editing
events involving the conversion of cytidine to uridine (C to U) and
represent RNA processing events by which several adenosine to inosine (A to I) occur as a result of hydrolytic
RNA and protein isoforms can be generated from a deamination at the C4 and C6 positions of the pyrimidine and
single genomic locus to increase the coding potential of purine bases, respectively, with the oxygen of water acting as
the human genome. the nucleophile. R: ribose.

ENCYCLOPEDIA OF LIFE SCIENCES & 2005, John Wiley & Sons, Ltd. www.els.net 1
 mRNA Editing

APOB gene
ATG CAA TAA

Transcription
Liver Intestine Splicing
APOB mRNA Polyadenylation

Nucleus
ATG CAA UAA ATG CAA (GIn) UAA
5⬘ (A)n 5⬘ (A)n
Nonedited mRNA Nonedited
mRNA C-to-U
editing

ATG UAA (Stop) UAA

(A)n

Cytoplasm
5⬘
Edited mRNA
Translation

apoB protein
1 4536 1 2152
Lipoprotein LDL receptor Lipoprotein
NH2 assembly binding
COOH NH2 assembly
COOH
(a) apoB-100 apoB-48

5⬘ efficiency Editing Mooring

element site sequence
Spacer
(b) ···CUUGAGACAUACGCGAUA C AAUUUGAUCAGUAUAUU···

Figure 2 Biosynthetic pathway for the tissue-specific production of apoB isoforms through site-specific C-to-U editing. (a) Organization
of the APOB gene, with the start (ATG), stop (TAA) and edited glutamine (CAA) codons indicated in exons 1, 26 and 29, respectively;
the cytidine residue in the nonedited intestinal transcript and the uridine residue in the edited mRNA, after the C-to-U modification, are
underlined. The structures of the apoB-100 and apoB-48 protein isoforms are shown below, with their principal functional domains and
corresponding amino acid lengths. (A)n, poly(A) tail. (b) The tripartite sequence motif required for the C-to-U editing of APOB transcripts
(efficiency element, spacer, mooring sequence), with the edited cytidine residue indicated in white on black lettering.

chylomicrons that deliver dietary lipids to the tissues. regulatory sequence surrounding the edited cytidine
By contrast, in the liver the unmodified (non- residue in APOB RNAs, with essential elements
edited) RNA encodes apoB-100, an integral 512-kDa referred to as the mooring sequence, enhancer and
component of lipoprotein complexes involved in the spacer region (Figure 2b; Backus and Smith, 1992;
transport of dietary cholesterol (Davidson and Shah et al., 1991).
Shelness, 2000). This C-to-U conversion emphasizes The biochemical characterization and purification
the potential power of RNA editing, because such a of enterocyte extracts showed that the editing of APOB
subtle RNA processing event can generate two distinct transcripts requires the actions of a multiprotein
isoforms of apoB protein that have unique roles in complex, referred to as the ‘editosome’, that contains
the synthesis, metabolism and transport of plasma a 27-kDa catalytic cytidine deaminase, apobec-1 and
lipoproteins. (See Apolipoprotein Gene Structure and several auxiliary proteins (Teng et al., 1993). Apobec-1
Function.) itself has a low affinity for APOB RNA in vitro and
Analyses of nuclear and cytoplasmic RNAs have cannot function in the absence of additional RNA-
shown that the C-to-U modification of APOB tran- binding factors (Anant et al., 1995). The auxiliary
scripts is a nuclear event that can be initiated at the protein apobec-1 complementation factor (ACF) has
time of polyadenylation and is essentially complete by been shown to bind specifically to the mooring
the time that the precursor mRNA (pre-mRNA) has sequence, interact with apobec-1 and promote the
been fully processed (Lau et al., 1991; Figure 2a). The C-to-U editing of APOB transcripts in the presence of
development of an in vitro assay system using apobec-1 (Mehta and Driscoll, 2002; Mehta et al.,
enterocyte postmicrosomal cytoplasmic extracts and 2000), thus defining the minimal holoenzyme required
radiolabeled APOB RNA substrates allowed for APOB modification.
investigators to determine that the mechanism respon-
sible for the C-to-U editing of APOB transcripts Neurofibromin 1
involves hydrolytic deamination at the C4 position of
cytidine (Figure 1; Hodges et al., 1991; Driscoll et al., In addition to APOB transcripts, a similar C-to-U
1989). This RNA modification requires a tripartite conversion has been observed in transcripts produced

2
 mRNA Editing

from neurofibromin 1 (NF1), a tumor suppressor gene deficiency, hyper-IgM syndrome, characterized by
involved in neurofibromatosis type I (Skuse et al., normal or increased concentrations of IgM with an
1996). Neurofibromatosis type I (also known as von absence of IgG, IgA and IgE, and a profound
Recklinghausen neurofibromatosis) is a complex here- susceptibility to bacterial infections (Notarangelo
ditary syndrome resulting in several abnormalities that et al., 1992). Whereas the role of apobec-1 is to edit
affect tissues derived from the embryonic neural crest, APOB RNAs by site-specific cytidine deamination, the
including cutaneous and subcutaneous neurofibromas function of AID is to potentiate changes in the
and malignancies of the central and peripheral nervous sequences of immunoglobulin genes themselves. Al-
system (Goldberg et al., 1996). The editing of NF1, though the mode of action for AID is completely
which was identified originally by using a computa- unknown, accumulating data have suggested that AID
tionally based strategy to find RNAs containing a high might act by either editing an RNA that encodes a key
degree of sequence conservation with the APOB editing protein (e.g. an endonuclease), thereby initiating the
site (Skuse et al., 1996), results in the conversion of a processes of CSR and SHM, or directly modifying the
genomically encoded arginine (CGA) codon to a stop target DNA, with the induced lesion triggering an
(UGA) codon, thereby generating a truncated isoform error-prone repair cascade (Martin et al., 2002).
of the NF1 protein. The physiological consequences of
this C-to-U conversion are not well understood, nor
are the roles that NF1 editing may have in the
pathogenesis of neurofibromatosis (Skuse and A-to-I Editing
Cappione, 1997). Notably, despite the sequence
similarity between the editing regions of APOB and The conversion of adenosine to inosine by RNA
NF1 transcripts, apobec-1 does not seem to be editing was first observed in yeast tRNAs (Grosjean
responsible for the modification of NF1 mRNAs et al., 1996; Holley, 1965) but has since been identified
(Skuse et al., 1996), indicating that different trans- in several viral RNA transcripts and cellular mRNA
acting factors may be involved in these two C-to-U species (Emeson and Singh, 2001). A-to-I editing
modifications. (See Neurofibromatosis: Type I.) is most often identified as an adenosine to guanosine
(A-to-G) discrepancy between genomic and comple-
mentary DNA (cDNA) sequences, owing to the
Immunoglobulin diversity similar base-pairing properties of inosine and
guanosine during cDNA synthesis. A-to-I conversion
The diversity of immunoglobulin expression required is catalyzed through hydrolytic deamination at the C6
for an effective immune response requires the gener- position of the adenine ring (Figure 1; Polson et al.,
ation of numerous immunoglobulin genes by distinct 1991) and requires an extended region of double-
cellular processes, including the recombination of stranded RNA in potential RNA substrates formed by
variable, diversity and joining segments (V(D)J intramolecular base-pairing interactions between exon
recombination), somatic hypermutation (SHM) and and intron sequences (Burns et al., 1997; Lomeli et al.,
class switch recombination (CSR; Honjo et al., 2002). 1994; Higuchi et al., 1993). (See RNA Secondary
After V(D)J gene rearrangement, mature B lympho- Structure Prediction; RNA-binding Proteins: Regulation
cytes undergo a switch of immunoglobulin isotype of mRNA Splicing, Export and Decay; tRNA.)
from immunoglobulin-m (IgM) to immunoglobulin-g A family of adenosine deaminases that act on RNA
(IgG), immunoglobulin-e (IgE) or immunoglobulin-a (ADARs) has been characterized extensively and is
(IgA) – a process that is mediated by CSR, which responsible for catalyzing the site-specific A-to-I
changes the immunoglobulin heavy chain constant conversion observed in numerous cellular mRNA
region (CH) gene from Cm to one of the other CH genes. transcripts (Melcher et al., 1996; O’Connell et al.,
SHM introduces vast numbers of point mutations in 1995; Hough and Bass, 1994; Kim et al., 1994).
the immunoglobulin variable region gene, giving rise Although only a few inosine-containing human tran-
to immunoglobulins with higher affinity. The process- scripts have been described so far, the wide distribu-
es of CSR and SHM depend on the activity of tion of ADAR protein expression and the relatively
activation-induced cytidine deaminase (AID), a B- high levels of inosine in mammalian brain mRNAs
cell-specific protein that has been proposed to function (Paul and Bass, 1998) have suggested that numerous
by RNA editing on the basis of its sequence similarity edited RNAs have yet to be identified. At least seven
to apobec-1 and other cytidine deaminases transcripts containing A-to-I editing events have been
(Muramatsu et al., 1999). (See Immunoglobulin Genes.) identified that can change the amino acid coding
Humans who are deficient in AID show a lack of potential in human mRNAs; notably, transcripts
CSR and severely reduced SHM (Revy et al., 2000), encoding the 2C subtype of the serotonin or
producing a rare autosomal recessive human immuno- 5-hydroxytryptamine (5-HT) receptor (5-HT2CR),

3
 mRNA Editing

the B-subunit of the a-amino-3-hydroxy-5-methyl- To address the role of Q/R editing and subsequent
isoxazole-4-propionate (AMPA) subtype of the iono- changes in the electrophysiological and ion-
tropic glutamate receptor (GluR-B) and the 5 subunit permeation properties of heteromeric AMPA channels,
and the 6 subunit of the kainate subtype of the Brusa et al. (1995) generated a genetically modified
ionotropic glutamate receptor (GluR-5 and GluR-6) mouse strain incapable of editing GluR-B transcripts
undergo A-to-I modifications that change the amino by selectively disrupting the RNA duplex region
acid coding potential of the mature mRNAs, thereby required for Q/R site editing. Animals heterozygous
producing protein products with altered functional for the targeted mutation were smaller than control
properties. A-to-I editing can also occur in noncoding littermates at birth and expressed AMPA receptors
regions of human mRNA transcripts (Morse et al., with the expected increase in Ca2þ permeability in
2002) and in introns (Rueter et al., 1995, 1999; Higuchi several brain regions. At roughly 2 weeks of age,
et al., 1993). however, the mutant animals rapidly developed a
severely compromised phenotype, characterized by
Glutamate-gated ion channels spontaneous and recurrent seizures, and by progres-
sively agitated states and running fits. All of the
The first example of A-to-I editing for mammalian heterozygous mutant animals died by postnatal day
mRNAs was observed in transcripts encoding 20, indicating that Q/R site editing is essential for
glutamate-gated ion channels in the central nervous normal brain function (Brusa et al., 1995).
system (CNS; Sommer et al., 1991). L-glutamate is the In addition to Q/R site modification in GluR-B
principal excitatory neurotransmitter in the brain and transcripts, a second RNA editing event converts an
activates cation-selective receptor channels involved in arginine (AGA) to a glycine (IGA) codon in mRNAs
fast synaptic neurotransmission and in the induction encoding the GluR-B, GluR-C and GluR-D AMPA
of long-term cellular changes associated with memory receptor subunits. Introduction of this edited glycine
acquisition and learning (Nicoll and Malenka, 1999; residue in heteromeric AMPA receptors results in an
Collingridge and Singer, 1990). Pharmacological and increase in the rate of recovery from receptor
physiological studies have identified three distinct desensitization and a tendency toward slower desensi-
subtypes of ionotropic glutamate receptor on the tization rates (Lomeli et al., 1994). These properties
basis of the selective agonists N-methyl-D-aspartate may be crucial for determining the excitability of a
(NMDA), AMPA and kainic acid. (See Glutamate single postsynaptic site in response to repeated
Receptors.) glutamate stimulation.
AMPA receptors Kainate receptors
The AMPA subtype of glutamate receptor is involved A second family of ionotropic glutamate receptors,
primarily in fast synaptic neurotransmission and is which are selectively responsive to kainic and domoic
expressed widely throughout the CNS (Hollmann and acid, can be generated by heteromeric assembly of the
Heinemann, 1994). Functional AMPA channels con- subunits GluR-5, GluR-6, GluR-7 (Bettler et al., 1992;
sist of homomeric or heteromeric combinations of four Sommer et al., 1992; Egebjerg et al., 1991) and KA-1 or
distinct subunits, GluR-A, GluR-B, GluR-C and KA-2 (Herb et al., 1992; Sakimura et al., 1990). Like
GluR-D (Boulter et al., 1990; Keinanen et al., 1990; AMPA receptors, kainate receptors are thought to be
Nakanishi et al., 1990; Sakimura et al., 1990). RNAs involved in fast synaptic neurotransmission and are
encoding the GluR-B subunit undergo a site-specific expressed widely in many regions of the brain
(Q/R site) modification that converts a genomically (Seeburg, 1993). The electrophysiological and ion-
encoded glutamine codon (CAG) to an arginine codon permeation properties of the GluR-5 and GluR-6
(CIG) in the second hydrophobic domain of the subunits can be regulated by RNA editing processes
protein that is presumed to line the ion channel pore. similar to those observed for AMPA receptor subunits
Heteromeric AMPA channels that contain the edited (Egebjerg and Heinemann, 1993; Kohler et al., 1993).
GluR-B subunit are impermeable to Ca2þ ions and In addition to Q/R editing sites in GluR-5 and
show a linear current–voltage (I–V ) relationship, GluR-6 transcripts, two additional adenosines are
whereas channels that lack or contain a nonedited selectively targeted for modification. Both of these
GluR-B subunit show a double-rectifying I–V rela- are located in the region encoding the first hydropho-
tionship and an increased Ca2þ conductance bic domain (TM1) of GluR-6, and editing leads to the
(Dingledine et al., 1992; Hume et al., 1991). substitution of a valine (ITT) for a genomically
The marked functional alterations observed for specified isoleucine (ATT) codon (I/V site) and the
heteromeric AMPA receptors containing the edited substitution of a cysteine (TIC) for a tyrosine (TAC)
GluR-B subunit have raised several issues regarding codon (Y/C site). These editing events provide the
the physiological relevance of this RNA modification. possibility of eight different edited variants of GluR-6

4
 mRNA Editing

subunits, all of which are expressed to a varying extent 5-HT2CR

in the CNS, although fully edited GluR-6 transcripts
represent the most abundantly expressed isoform in Serotonin or 5-HT is a monoaminergic neurotrans-
the adult nervous system (Ruano et al., 1995; Kohler mitter that modulates many different behaviors,
et al., 1993). including sleep, appetite, pain perception, locomotion,
Editing at the GluR-6 Q/R site can regulate calcium thermoregulation and sexual behavior. The several
permeability through kainate receptor channels; how- actions of 5-HT are mediated by 14 distinct subtypes
ever, the amino acid residue present at this position of 5-HT receptor that differ in their tissue localization,
determines calcium permeability in a manner opposite binding affinity for 5-HT and coupling to distinct
to that observed for AMPA receptors (Kohler et al., intracellular signaling pathways (Hoyer et al., 1994).
1993). The physiological relevance of these functional The 5-HT2 family of receptors has three members,
alterations has yet to be determined, however, because 5-HT2AR, 5-HT2BR and 5-HT2CR, that activate
mutant mice capable of expressing only the edited phospholipase C to increase intracellular
GluR-5(R) isoform show no developmental abnor- concentrations of inositol phosphates and
malities or deficits in many behavioral models (Sailer diacylglycerol (Figure 3). The 5-HT2A and 5-HT2C
et al., 1999). receptors are thought to be involved in many human

W V L L G I L H V L F H V A N R L N V M NH2
Q
CD I S V S P V A A I
V NN T
T F V T
VWP L V C
D CN
K
F R GGD S T N F I
D
Y P K V K S Q
R E L E K
F Y
P L Y E N C L
D I L D D L M
G A C R P V E
V L P L N S K Extracellular
WNQ L
I G L N L L
S L A P M P L S I WV
P I P V L V F I T N I
I V I L L V V D L S F S G I WV F V
M I I I M L V G S F L V S V F A V P F F Y G I
I S A T I S I G L P I F I MW C G S C V
G I T I A D A L P N I
L I NG L HM V W M I T V F
MS L A K I A I I V F F T Y V L
M I V Y F L S I A C Y T I V L G F L
A I M V
V A
S N L K L K N Intracellular
M T D T T S K
E A R R I A I
K N Y S Y K Y R R A F S N Y L R CNY
K L H V
R
V N L N E K
A F R A I N Q R V P P K K E V
R Q
R G T MQ A V R P I
I/V/M S A R A
M L P R R E K K K R
H L T A L S G
R E L R R E L N V N I
R
N/S/G I/V H A Y
P G
H N ND S A K E I V P E N T HR
Q E
T P
RNA editing E D
G
Q I E MQ V E N L E L P V
region E N N
P P P
P N S
A HOOC V S S I R E S V V S
G S
L E N
S E E
(a) L D F L K C C K R N T A

A GA
B EC D G C
C U
U C A
A A U U G A U C A UAUU
A UCGGU UGU A GC A U A C G U A A U C C U A U U G A G C A U A G C C G C A A U U C C G C C A A UG AGA UUGC G
UGGU C A A C A UG U G U A U G U A U U A G G A U A A C U C G U A U C G G C G U U A A G C C G G UU U A U UUUA ACG U
G G C G C C C G C A A U C GGUU
C A A A U U A AU U A
U C A A U U G Exon
G C CG
A A UA Intron
A A
(b) A A
U

Figure 3 RNA editing of 5-HT2CR transcripts. (a) Representation of the amino acid sequence and predicted topology of the
human 2C subtype of serotonin receptor. The positions of amino acid alterations in the second intracellular loop of the receptor, which
result from A-to-I editing events, are indicated, along with possible amino acid residues that result from permutations of editing at five
distinct sites. (b) Secondary structure of the region of major editing modifications in the pre-mRNA transcript encoding the 5-HT2CR,
as predicted by RNA folding algorithms (Zuker, 1989); the positions of the five edited adenosine residues (labeled A–E) and the exon–intron
boundary are indicated. The codons altered by RNA editing are emboldened and shaded to match the amino acid positions shown in (a).

5
 mRNA Editing

psychiatric and behavioral disorders, including psy- genes (Venter et al., 2001) – far less than the number
chotic depression, dysthymia, obsessive–compulsive previously anticipated to support the highly sophisti-
disorder, anxiety and schizophrenia (Dubovsky and cated biological functions of human beings. Despite
Thomas, 1995; Pandey et al., 1995; Teitler and this apparent paucity of genes, however, the main
Herrick-Davis, 1994; Julius, 1991). Mutant mice in focus of the postgenomic era will most likely turn
which expression of the 5-HT2CR has been selectively toward describing and functionally characterizing
ablated show leptin-independent hyperphagia and the full complement of proteins (i.e. the proteome)
type 2 diabetes (Nonogaki et al., 1998), death from that are expressed by an organism. The number of
spontaneous and audiogenic seizures (Applegate and proteins in the proteome is by no means equivalent to
Tecott, 1998; Brennan et al., 1997; Tecott et al., 1995) the number of genes itself, but is also dependent on
and learning deficits resulting from hippocampal numerous cellular strategies that function to increase
dysfunction (Tecott et al., 1998). protein diversity, such as DNA recombination,
RNA transcripts encoding the 5-HT2CR undergo multiple transcription start sites, alternative pre-
up to five A-to-I editing events (at sites A–E), which mRNA splicing, RNA editing and posttranslational
predict alterations in the identity of three amino acids protein modification. (See RNA Processing; Splicing of
in the second intracellular loop of the receptor pre-mRNA.)
(Figure 3a; Niswender et al., 1999; Burns et al., Because most editing events in human mRNAs
1997) – a region involved in coupling to G proteins have been serendipitously identified on the basis of
(Pin et al., 1994; Wong et al., 1990). Sequence analysis discrepancies between genomic and cDNA sequences,
of cDNAs isolated from dissected rat and human it is not necessarily surprising that only a few edited
brains has shown the region-specific expression of as transcripts have been described so far. With the com-
many as 12 principal 5-HT2C receptor isoforms pletion of the Human Genome Project, however, our
encoded by 18 distinct RNA species (Niswender et al., ability to identify additional edited RNAs, by using
1999; Burns et al., 1997), suggesting that differentially bioinformatic, genetic and biochemical approaches,
edited 5-HT2C receptors may have distinct biological should accelerate rapidly. By using such genomic
functions in the regions in which they are expressed. As sequence information, many A-to-I modified human
with all identified A-to-I editing substrates, the brain transcripts have been identified already on the
specificity of 5-HT2CR editing results from an RNA basis of the presence of inosine in the mature mRNAs
duplex region in the 5-HT2CR pre-mRNA formed by (Morse et al., 2002). Unexpectedly, these newly identi-
base-pairing interactions between exon and intron fied RNA species contain editing events in noncoding
sequences (Figure 3b; Burns et al., 1997). RNAs or noncoding regions of mRNAs, suggesting
Functional comparisons of the nonedited 5-HT2CR that RNA editing may influence many aspects of
(containing isoleucine, asparagine and isoleucine at cellular RNA function. (See 30 UTRs and Regulation;
amino acids 157, 159 and 161 respectively) and the 50 UTRs and Regulation; Noncoding RNAs: A Regulatory
edited 5-HT2CR isoform (containing valine, glycine Role?)
and valine at the analogous positions) identified a Although seemingly subtle single-base modifica-
40-fold decrease in serotonergic potency to stimulate tions in open reading frames can generate multiple
the hydrolysis of phosphoinositides; this decrease in protein isoforms with distinct biological activities
serotonin potency results from a reduced G-protein (Gott and Emeson, 2000; Rueter and Emeson, 1998),
coupling efficiency induced by the introduction of editing events in introns can affect splicing patterns
novel amino acids into the second intracellular loop of (Rueter et al., 1999; Wissinger et al., 1991) and editing
the receptor (Niswender et al., 1999). Other studies events in noncoding RNAs or in the 50 and 30
have indicated that alterations in 5-HT2CR editing are untranslated regions of mRNAs (Morse et al., 2002;
present in individuals diagnosed with schizophrenia Keller et al., 1999) may modulate processes such as
(Sodhi et al., 2001), as well as in suicide victims with a RNA stability, localization and translation efficiency.
history of major depression (Gurevich et al., 2002; The ultimate biological relevance of editing for
Niswender et al., 2001), suggesting that editing of the human mRNAs resides in the specific RNA transcripts
5-HT2CR may be involved in neuropsychiatric that are modified by these processes, and it is
disorders and in the maintenance of appropriate anticipated that the completed human genome se-
serotonergic signaling. quence will not only serve to identify additional edited
transcripts, but also to determine the full array of
cellular processes in which RNA editing has a role.
Perspectives and Future Directions (See mRNA Localization: Mechanisms; mRNA Stability
and the Control of Gene Expression; mRNA Turnover;
Analysis of the completed human genome sequence RNA Editing and Human Disorders; RNA Processing and
has suggested that it may contain as few as 30 000 Human Disorders.)