Molecular identification of the long isoform of the human neuropeptide Y Y5 receptor and pharmacological comparison with the short Y5 receptor isoform.

Abstract

The neuropeptide Y Y5 receptor gene generates two splice variants, referred to here as Y5(L) (long isoform) and Y5(S) (short isoform). Y5(L) mRNA differs from Y5(S) mRNA in its 5' end, generating a putative open reading frame with 30 additional nucleotides upstream of the initiator AUG compared with the Y5(S) mRNA. The purpose of the present work was to investigate the existence of the Y5(L) mRNA. The authenticity of this transcript was confirmed by isolating part of its 5' untranslated region through 5' rapid amplification of cDNA ends and analysing its tissue distribution. To study the initiation of translation on Y5(L) mRNA, we cloned the Y5(L) cDNA and two Y5(L) cDNA mutants lacking the first or the second putative initiation start codon. Transient expression of the three plasmids in COS-7 cells and saturation binding experiments using (125)I-labelled polypeptide YY (PYY) as a ligand showed that initiation of translation on Y5(L) mRNA could start at the first AUG, giving rise to a Y5(L) receptor with an N-terminal 10-amino-acid extension when compared with the Y5(S) receptor. The human Y5(L) and Y5(S) receptor isoforms displayed similar affinity constants (1.3 nM and 1.5 nM respectively). [(125)I]PYY binding to COS-7 cells expressing either the Y5(L) or the Y5(S) isoform was inhibited with the same rank order of potency by a selection of six chemically diverse compounds: PYY>neuropeptide Y>pancreatic polypeptide>CGP71683A>Synaptic 34>Banyu 6. Comparison of the tissue distribution of Y5(L) and Y5(S) mRNAs, as determined by reverse transcription-PCR analysis, indicated that expression of Y5(L) mRNA occurs in a tissue-specific manner. Finally, we have shown that the two AUG triplets contained in the 5' untranslated region of Y5(L) mRNA did not affect receptor expression.

Full text

Page 1

Biochem. J. (2003) 369, 667–673 (Printed in Great Britain)
667
Molecular identification of the long isoform of the human neuropeptide Y
Y5 receptor and pharmacological comparison with the short Y5 receptor
isoform
Marianne RODRIGUEZ, Vale?rie AUDINOT, Sandra DROMAINT, Christelle MACIA, Ve?ronique LAMAMY, Philippe BEAUVERGER1,
Herve? RIQUE2, Je?ro ? me IMBERT3, Jean Paul NICOLAS4, Jean A. BOUTIN5and Jean Pierre GALIZZI
Institut de Recherches Servier, Division de Pharmacologie Mole?culaire et Cellulaire, 125 chemin de Ronde, 78 290 Croissy sur Seine, France
The neuropeptide Y Y5 receptor gene generates two splice
variants, referred to here as Y5L(long isoform) and Y5S(short
isoform). Y5LmRNA differs from Y5SmRNA in its 5? end,
generating a putative open reading frame with 30 additional
nucleotides upstream of the initiator AUG compared with the
Y5SmRNA. The purpose of the present work was to investigate
the existence of the Y5LmRNA. The authenticity of this
transcript was confirmed by isolating part of its 5? untranslated
region through 5? rapid amplification of cDNA ends and
analysing its tissue distribution. To study the initiation of
translation on Y5LmRNA, we cloned the Y5LcDNA and two
Y5LcDNA mutants lacking the first or the second putative
initiation start codon. Transient expression of the three plasmids
in COS-7 cells and saturation binding experiments using ???I-
labelled polypeptide YY (PYY)as a ligand showed that initiation
of translation on Y5LmRNA could start at the first AUG, giving
risetoaY5LreceptorwithanN-terminal10-amino-acidextension
INTRODUCTION
Neuropeptide Y (NPY) is a 36-amino-acid peptide belonging to
the pancreatic polypeptide family, which also includes poly-
peptide YY (PYY) and pancreatic polypeptide (PP) [1]. NPY is
one of the most abundant and widely distributed neuro-
transmitters in both the central and peripheral nervous systems
[2]. Chronic administration of NPY into the lateral ventricle or
direct injection into the paraventricular nucleus of the rat
increases food intake and lead to obesity [3,4]. In addition to its
orexigenic action, NPY modulates numerous physiological pro-
cesses,includinganxiety,anti-convulsantactivity,bloodpressure,
memory retention, circadian rhythms and gastrointestinal func-
tion [5–10]. These diverse physiological actions are mediated
through distinct NPY receptor subtypes, all of which are coupled
to G-proteins, and their activation leads to the inhibition of
adenylate cyclase. To date, five NPY receptors, designated Y1,
Y2, Y4, Y5 and Y6, have been cloned from human and rodent
tissues (reviewed by Blomqvist and Herzog [11]). Among these,
the NPY Y1 and NPY Y5 receptors are regarded as putative
Abbreviations used: CRE, cAMP response element; MSH, melanocyte-stimulating hormone; NPY, neuropeptide Y; PP, pancreatic polypeptide; PYY,
polypeptide YY; RACE, rapid amplification of cDNA ends; RT-PCR, reverse transcription–PCR; UTR, untranslated region; Y5Land Y5S, long and short
isoforms respectively of the neuropeptide Y Y5 receptor.
1Present address: Sanofi-Synthelabo Recherche, Departement Cardiovasculaire Thrombose 1, Avenue Pierre Brossolette, 91385 Chilly-Mazarin
Cedex, France.
2Present address: Aventis Pharma S.A., Centre de Recherche de Paris Etablissement de Vitry 13, quai Jules Guesdes, F 94400 Vitry-sur-Seine,
France.
3Present address: Biogema, 24 Avenue des Landais, 63170 Aubie ? re, France.
4Present address: EntoMed SA, Rue Tobias Stimmer, 67400 Illkirch, France.
5To whom correspondence should be addressed (e-mail jean.boutin?fr.netgrs.com).
when compared with the Y5Sreceptor. The human Y5Land Y5S
receptor isoforms displayed similar affinity constants (1.3 nM
and 1.5 nM respectively). [???I]PYY binding to COS-7 cells
expressing either the Y5Lor the Y5Sisoform was inhibited with
the same rank order of potency by a selection of six chemically
diversecompounds:PYY?neuropeptide
polypeptide?CGP71683A?Synaptic 34?Banyu 6. Com-
parison of the tissue distribution of Y5Land Y5SmRNAs, as
determined by reverse transcription–PCR analysis, indicated
that expression of Y5LmRNA occurs in a tissue-specific manner.
Finally, we have shown that the two AUG triplets contained in
the 5? untranslated region of Y5LmRNA did not affect receptor
expression.
Y?pancreatic
Key words: initiation codon context, NPY receptor, obesity,
translation, 5? untranslated region.
receptor subtypes that mediate the appetite-stimulating action of
NPY [12–17].
Initial cloning of the Y5 receptor in humans identified two
alternatively spliced Y5 mRNA transcripts [15,18,19]. Sequence
comparison of the corresponding cDNAs indicated divergence in
their 5? untranslated regions (UTRs) and 5? end coding regions.
Gerald et al. [15] reported a cDNA encoding a predicted 455-
amino-acid receptor, whereas the Y5 cDNA described by Hu et
al. [18] encoded a shorter 445-amino-acid version of the receptor
lacking the 10 N-terminal amino acids. We thus designated the
longer and the shorter forms of the receptor Y5Land Y5S
respectively. Parker and Xia [20] showed evidence for extensive
alternative splicing in the 5? UTR of the Y5 receptor gene. Five
human NPY Y5 receptor cDNAs splice variants, all of which
encoded the short form of the Y5 receptor, were isolated by
performing 5? rapid amplification of cDNA ends (RACE) from
brain poly(A+) RNA. Moreover, while we and others [21,22]
reported the cloning and the expression of the Y5Lform, Parker
et al. [23] explained their unsuccessful attempts to express the
Y5Lreceptor by suggesting that nucleotides 1–55 of the sequence
? 2003 Biochemical Society

Page 2

668
M. Rodriguez and others
reported by Gerald et al. [15] arise from the NPY Y5 receptor
gene and are not actually found in human Y5 cDNA transcripts.
In view of this controversy, the purpose of the present work
was to investigate the existence of the Y5LmRNA. The auth-
enticity of this transcript was indeed confirmed by isolating part
of its 5? UTR through 5? RACE and analysing its tissue
distribution. Moreover, we provide evidence that initiation of
translation on the Y5LmRNA can give rise to a functional
receptor with an N-terminal 10-amino-acid extension when
compared with the Y5Sreceptor, and that the two AUG triplets
contained in the 5? UTR did not affect receptor expression.
Finally, we show that the Y5Land Y5Sreceptor isoforms display
similar pharmacological profiles.
MATERIALS AND METHODS
5? RACE
PCR was carried out with Clontech Human Hypothalamus
Marathon-Ready cDNA as a template, according to the manu-
facturer’s instructions (Clontech, Palo Alto, CA, U.S.A.). In the
first PCR reaction, cDNA was amplified with the abridged
anchor primer AP1 and a reverse primer based on a region
common to the Y5Sand Y5Lsequences (nucleotides 354–383;
GenBank accession no. U56079). The resulting PCR products
were then submitted to Southern blotting with internal specific
oligonucleotide probes based on unique regions of the Y5L(OP1;
nucleotides 18–40; accession no. U56079) and Y5S(OP2; nucleo-
tides 39–61; accession no. U94320) sequences. For nested PCR,
the primary PCR product was subjected to 25 cycles of ampli-
fication using the abridged anchor primer AP2 and a reverse
primer based on a region common to the Y5Sand Y5Lsequences
(nucleotides 296–322; accession no. U56079). The RACE
products were then subcloned into pT-Adv (Clontech).
Analysis of expression by reverse transcription–PCR (RT-PCR)
RNA from human hypothalamus was obtained from Analytical
Biological Services (Wilmington, DE, U.S.A.), and RNA from
human hippocampus and caudate nucleus was obtained
from Clontech. RNAs were reverse transcribed using oligo-
(dT)??–??and Superscript II reverse transcriptase (Invitrogen).
The first-strand cDNA (corresponding to 1 µg of total RNA)
wasamplified using a programconsisting of30 cycles of94 ?C for
1 min, 55 ?C for 1 min and 72 ?C for 3 min, with pre- and
post-incubation steps of 94 ?C for 1 min and 72 ?C for 5 min
respectively. PCR amplification utilized forward primers based
on unique regions of the Y5L(nucleotides 18–40; accession
no. U56079) and Y5S(nucleotides 39–61; accession no. U94320)
sequences and a reverse oligonucleotide primer (SD5) based on a
region common to the Y5Sand Y5Lsequences (nucleotides 364–
383; accession no. U56079). PCR products were separated by
agarose (1%, w?v) gel electrophoresis and transferred to a
HybondN+membrane (Amersham PharmaciaBiotech). Hybrid-
ization was performed at 42 ?C with internal specific Y5 ??P-
labelled oligonucleotide probe (nucleotides 56–77; accession
no. U56079).
Construction of recombinant plasmids
Poly(A+) RNA from human brain caudate nucleus (Clontech)
was reverse transcribed with oligo(dT)??–??using Superscript II
reverse transcriptase. First-strand cDNA (corresponding to 1 µg
of total RNA) was amplified using a program consisting of 35
cycles of 94 ?C for 1 min, 55 ?C for 1 min and 72 ?C for 3 min,
with pre- and post-incubation steps of 94 ?C for 1 min and 72 ?C
for7 minrespectively.PCRamplificationutilizedoligonucleotide
primersbasedonGenBankentriesforeitherhumanY5L(forward
18–40, reverse 1377–1399; accession no. U56079) or human Y5S
(forward 51–73, reverse 1377–1399; accession no. U56079).
The expected PCR fragments were isolated and ligated into the
Okayama and Berg [24] expression vector pSR. The recombinant
plasmids pY5Sand pY5Lwere sequenced on both strands by
automated sequencing.
To study the initiation of translation on Y5LmRNA, two Y5L
cDNA mutants lacking the first or the second putative initiation
start codon were constructed by PCR cloning. The mutant
MATG1 had the first ATG codon (nucleotides 26–28; accession
no. U56079) mutated to GTG. The mutant MATG2 had the
second ATG codon (base 56–58; accession no. U56079) mutated
to GTG. The template used in PCR was the plasmid pY5L
described above. The expected PCR fragments were isolated and
ligated into the Okayama and Berg [24] expression vector pSR.
After subcloning, the presence of the indicated mutations was
confirmed by automated DNA sequencing.
To study the effects of upstream AUG codons on Y5LmRNA
translation, the Y5Lcoding region flanked by 50 bp from the 5?
UTR was amplified from human brain caudate nucleus cDNA
by PCR with nucleotide primers (5? GAATAGATTAATTTA-
AAGTAGTCA 3?) and (5? GAATTATTACATATGAAGAC-
AGT 3?). The expected PCR fragment was isolated and ligated
intotheexpressionvectorpSR.Therecombinantplasmid pUY5L
was sequenced on both strands by automated sequencing. The
mutant pUY5L1 had the first ATG codon (see Figure 1B)
mutated to GTG. The mutant pUY5L2 had the second ATG
codon (see Figure 1B) mutated to GTG. The templates used in
PCR were the plasmids pUY5Land pUY5L1 described above.
The expected PCR fragments were isolated and ligated into
the expression vector pSR. After subcloning, the presence of the
indicated mutations was confirmed by automated DNA
sequencing.
Receptor expression
COS-7 cells grown in Dulbecco’s modified Eagle’s medium
containing10%(v?v)fetalcalfserumwereseededat15?10? cells
per 225 ml culture flask, and transfected 24 h later with 25 µg of
plasmid, using AMINETM as described by the manu-
facturer (Life Technologies). Cells were scraped 48 h after
transfection into lysis buffer (20 mM Tris?HCl, pH 7.7, con-
taining5 mMEDTA),andmembraneswerepreparedforbinding
assays.
Preparation of cell membranes
The suspension was homogenized using a Kinematica polytron
and then centrifuged at 43000 g (30 min, 4 ?C). The resulting
pellet was resuspended in binding buffer (see below) without
BSA at a concentration of 5 mg?ml. Aliquots of membrane
preparation were stored at ?80 ?C until use.
[125I]PYY binding assay
Membranes (50 µg?ml) were incubated for 90 min at 30 ?C in
binding buffer (20 mM Hepes, pH 7.4, containing 10 mM NaCl,
0.22 mM KH?PO?, 1.26 mM CaCl?, 0.81 mM MgSO?and 0.1%
BSA) in a final volume of 500 µl containing 0.045 nM [???I]PYY
and the tested drug for displacement experiments. Non-specific
binding was defined using 1 µM NPY. For saturation experi-
ments, isotopic dilutions were performed. Briefly, a range of
concentrations starting from 0.1–10 nM [???I]PYY?PYY with a
fixed PYY?[???I]PYY ratio of 10 was tested. Incubations were
stopped by rapid filtration through GF?C unifilters presoaked in
? 2003 Biochemical Society

Page 3

669
Pharmacology of short and long neuropeptide Y Y5 receptor isoforms
Figure 1Analysis of the 5? end region of Y5LcDNA
(A) PCR product analysis. The 5? RACE experiment was performed with a human brain cDNA library, as described in the Materials and methods section. The PCR products were run on a 1%
(w/v) agarose gel, transferred to a Hybond N+membrane, and hybridized with the OP1 oligonucleotide probe, based on a unique region of the Y5Lsequence, or the OP2 oligonucleotide probe,
based on a region common to four out of the five Y5Stranscripts. (B) Nucleotide sequence of the 5? end region of Y5LcDNA. The 5? UTR appears in lower-case and the coding region in upper-
case letters. The putative initiation start codons (ATG) are underlined. The ATG codons located in the 5? UTR are indicated in bold.
0.1% polyethyleneimine, followed by three successive washes
with 50 mM Tris?HCl, pH 7.4.
cAMP response element (CRE)–luciferase reporter assay
The p∆MC16-Luc plasmid (provided by Dr J. Bockaert, UPR
CNRS 9023, Montpellier, France), which contains the luciferase
coding sequence driven by a thymidine kinase promoter flanked
by16cAMPresponseelements(CREs),andthepcDOR8plasmid
(provided by P. Morgan, Rowett Research Institute, Aberdeen,
Scotland), encoding the expression of the ovine MC5 melano-
cortin receptor, have been described previously [25,26]. This
method is an adaptation of the method described by Conway et
al.[27]forthecharacterizationofthehumanmelatoninreceptors.
HEK cells, cultured to ?80% confluence, were transfected with
AMINE PlusTM (Life Technologies) using 20 ng of
pcDOR8, 500 ng of p∆MC16-Luc and 1 µg of either pY5Lor
pY5Sper 10? cells. At 24 h after transfection, cells were detached
by agitation, centrifuged (12000 g, 10 min) and recovered at
10? cells?ml in Phenol Red-free Dulbecco’s modified Eagle’s
medium supplemented with 2% (v?v) fetal calf serum. They were
then seeded as 50 µl aliquots into white 96-well tissue culture
plates. The cells were treated first with [Nle?--Phe?]α-
melanocyte-stimulating hormone (α-MSH; 10 nM), and then
with increasing concentrations of NPY (10 pM–1 µM). After a
20 h incubation period, luciferase activity was measured by
adding 100 µl per well of Luclite buffer (Packard, Meriden, CT,
U.S.A.) following by a 30 min incubation in the dark at room
temperature, and fluorescence was read using a TopCount plate
reader (Packard).
Chemicals
Allpeptide referencecompounds wereobtained from NeoSystem
(Strasbourg, France) or Bachem (Basel, Switzerland). Banyu 6 is
compound 6 described by Fukami et al. [28], and Synaptic 34
is compound 34 described by Islam et al. [29].
Data analysis
Binding data were generated as duplicate values within each
experiment, which were repeated independently at least three
times. Saturation analysis was analysed using the program
PRISM (GraphPad Software Inc., San Diego, CA, U.S.A.) to
yield KD(the dissociation constant of the radioligand) and Bmax
(the maximal number of binding sites). Displacement curve
fittings were generated by non-linear regression to yield IC??
values (concentration of compound giving 50% inhibition of
[???I]PYY binding). Inhibition constants (Ki) were calculated
according to the Cheng–Prusoff equation: Ki?IC???(1?L?KD),
where L is the concentration of [???I]PYY. Results are expressed
as mean pKi(?logKi).
RESULTS AND DISCUSSION
Identification of the Y5Ltranscript
In order to confirm the authenticity of the Y5Ltranscript, 5?
RACE was conducted using human hippocampus Marathon-
Ready cDNA as a template. PCR was performed with the
forward adapter AP1 primer and a reverse primer based on a
regioncommontotheY5SandY5Lsequences.ThePCRproducts
were then characterized by Southern blotting with two oligo-
nucleotide probes: OP1, based on a unique region of the Y5L
sequence, and OP2, based on a region common to four out of the
five Y5Stranscripts described by Parker and Xia [20]. As shown
in Figure 1(A), bands ranging in size between 400 and 920 bp
were detected with the OP1 probe, whereas two bands with
approximate sizes of 420 bp and 750 bp were found with the OP2
probe. To further characterize the 5? UTR of the Y5LmRNA,
the resulting PCR products were cloned. This initial attempt was
unsuccessful, as no clone was detected by hybridization with the
? 2003 Biochemical Society

Page 4

670
M. Rodriguez and others
Figure 2Distribution of Y5Land Y5SmRNAs in human tissues
cDNAs from different brain regions (A) and peripheral tissues (B) were used as templates for
PCR amplification. Two sets of primers, OP1-SD5 and OP2-SD5, were designed to amplify
fragments that are unique to the Y5Land Y5Ssequences respectively. The PCR products were
run on a 1% (w/v) agarose gel, transferred to a Hybond N+membrane, and hybridized with
aspecific32P-labelled oligonucleotide probe. RT – indicates control experiments lacking reverse
transcriptase.
OP1 probe. Therefore a second round of PCR was performed
using nested primers. The resulting PCR products were cloned
and three positively hybridizing clones were isolated. The se-
quence of the clone with the longest 5? sequence (Figure 1B)
extended 80 bp upstream from the translation initiation codon
reported by Hu et al. [18] and was in complete agreement with
the cDNA described in the patent literature by Gerald et al. [15].
Despite our extensive screening, we were not able to obtain a full-
length 5? UTR for the Y5LmRNA, suggesting that there are
difficulties associated with cloning this region. This might explain
why others have failed in their attempts to identify the Y5L
cDNA [19,23].
To provide further evidence that the Y5LcDNA was not an
artifact of the 5? RACE procedure, RT-PCR was performed
from human caudate nucleus poly(A)+RNA using OP1 as the
forward primer. A PCR fragment of the expected size was
obtained and hybridized to an internal oligoprobe (results not
shown). This result is consistent with the sequence data obtained
earlier by 5? RACE and confirms the existence of the Y5LmRNA
transcript. Moreover, our data are supported by the first Y5
cloning report published by Gerald et al. [15] in which they
described the Y5Lsequences isolated from human hippocampus
and rat hypothalamus cDNA libraries. It therefore appears that
Figure 3Initiation of translation on Y5LmRNA
Schematic representations of Y5LcDNA constructs are shown. The first or the second putative
initiation start codon (ATG) was mutated to GTG as indicated. The three expression vectors
pY5L, MATG1 and MATG2 were transiently expressed in COS-7 cells. Y5 receptor expression
(Bmax) was measured by radioligand binding studies carried out on membrane preparations
using [125I]PYY, as described in the Materials and methods section.
theY5LmRNA transcriptretains intronic sequence inthe mature
mRNA.
Pattern of expression of Y5LmRNA in human brain and
peripheral tissues
We next examined the distribution of hY5Land hY5SmRNAs in
severalhumanbrain regions using RT-PCR. Two setsof primers,
OP1-SD5 and OP2-SD5, were designed to amplify fragments
that are unique to the Y5Land Y5Ssequences respectively.
Southern blotting with an internal oligonucleotide probe indi-
cated the authenticity of the amplicons. As shown in Figure 2(A),
the Y5Land Y5Stranscripts displayed the same expression
pattern. No signal was observed when reverse transcriptase was
omitted from the first-strand cDNA conversion, which suggests
that the signals observed were not due to any genomic DNA
contaminating the RNA. Surprisingly, only low levels of Y5
mRNA were detected in the hypothalamus. This result contrasts
with previous in situ hybridization and PCR data that have
shown relatively high levels of Y5 mRNA in this brain area [30].
These differences might be explained by the fact that the
hypothalamus preparation originated from a single patient. High
levels of Y5 mRNA were detected in the caudate nucleus and
hippocampus. These findings are consistent with data obtained
in Northern blotting and in situ hybridization studies [30,31].
Woldbye et al. [32] suggested that Y5 receptors mediate the
anticonvulsant action of intraventricularly injected NPY on
motor convulsions and EEG seizures, which are induced by
systemic administration of kainic acid in rats. In addition, Kopp
et al. [33] demonstrated that seizure activity regulates gene
expression for Y5 receptors in the limbic system. It would thus be
of interest to comprehensively map and compare, at a cellular
level, the differential distributions of Y5Land Y5SmRNA
variants in limbic structures.
To investigate expression in peripheral tissues, we performed
PCR with a set of normalized human tissue cDNAs. As shown
in Figure 2(B), Y5Sand Y5LmRNAs displayed different ex-
pression patterns. For Y5S, the highest level of expression was
found in testis. Colon, ovaries, prostate and spleen showed inter-
mediate levels, while a very low level of expression was found
in small intestine. In this experiment, no expression was de-
tected in leucocytes. In contrast with Y5S, Y5Ltranscripts were
expressed at similar levels in colon, spleen and testis. Lower
levels were found in leucocytes and prostate, whereas no ex-
? 2003 Biochemical Society

Page 5

671
Pharmacology of short and long neuropeptide Y Y5 receptor isoforms
Figure 4
activity
Inhibition by NPY of [Nle4-D-Phe7]α-MSH-stimulated luciferase
HEK cells were co-transfected with plasmids pcDOR8 p∆MC16-Luc and either pY5Lor pY5S.
Transfected cells were then exposed to 10 nM [Nle4-D-Phe7]α-MSH and increasing concen-
trations of NPY (10 pM–1 µM) for 20 h. Data represent normalized values from experiments
carried out in duplicate.
pression was detected in the small intestine. These data suggest
that the expression of Y5 mRNA isoforms occurs in a tissue-
specific manner. Moreover, Y5 receptor gene expression may be
Figure 5 Inhibition of [125I]PYY binding to COS cell membranes transiently expressing Y5Lor Y5S
(A) Dose-dependent inhibition of specific binding of [125I]PYY in the presence of NPY and CGP71683A. Values shown are from a representative experiment performed in duplicate and repeated
at least three times. (B) Potency of compounds to inhibit specific binding of [125I]PYY. Values are expressed as mean pKi(?logKi) from at least three independent experiments, each performed
in duplicate. r and h denote rat and human respectively.
under the control of multiple promoters that are activated in a
tissue-specific manner.
Initiation of translation on Y5LmRNA
According to a widely accepted model proposed by Kozak [34],
recognition by eukaryotic ribosomes of the translation initiation
codon generally proceeds by a scanning mechanism in which the
40 S ribosomal subunit migrates from the capped mRNA 5? end
to the first AUG triplet that is found in a favourable context. The
efficiency of initiation from an AUG codon is increased if it lies
in an optimal sequence context, which in higher eukaryotes is
CA?GCCAUGG, with purines being critical at positions ?3 and
?4 [35]. Surprisingly, the sequence ACAUAUGU surrounding
the AUG codon in Y5LmRNA assigned by Gerald et al. [15] as
the initiator codon does not conform to the typical Kozak
consensus sequence. The in-frame AUG located 30 nucleotides
downstream lies in a much better sequence context, i.e.
UAAUAUGG, and would give rise to the protein described by
Hu et al. [18]. In order to investigate the initiation of translation
on Y5LmRNA, wild-type Y5LcDNA and two Y5LcDNA
mutants, MATG1 and MATG2, respectively lacking the first
and the second putative initiation start codon, were cloned and
transiently expressed in COS-7 cells. The levels of expression
were measured by saturation binding experiments carried out on
membrane preparations using [???I]PYY.
As shown in Figure 3, the three constructs exhibited similar
levels of expression. If initiation was limited to the second AUG
? 2003 Biochemical Society