Báo cáo khoa học: Cloning of type 1 cannabinoid receptor in Rana esculenta reveals differences between genomic sequence and cDNA pot

Alignments of FCNR1 with those of other vertebrates revealed amino acid identity ranging from 61.9% to 88.1%; critical domains for CNR1 functionality were conserved in the frog.. As nucl

esculenta reveals differences between genomic sequence and cDNA

Rosaria Meccariello1, Rosanna Chianese2, Gilda Cobellis2, Riccardo Pierantoni2 and Silvia Fasano2

1 Dipartimento di Studi delle Istituzioni e dei Sistemi Territoriali, Universita` di Napoli ‘Parthenope’, Naples, Italy

2 Dipartimento di Medicina Sperimentale, Sez ‘F Bottazzi’, II Universita` di Napoli, Naples, Italy

Cannabinoid receptors (CNRs) bind D9

-tetrahydrocan-nabinol, the major active constituent of the marijuana

plant, Cannabis sativa, and some endogenous lipidic

mediators collectively termed ‘endocannabinoids’ [1,2]

The best known endogenous ligands for CNRs are

anandamide (arachidonoylethanolamide, AEA),

2-arach-idonoylglycerol, the noladin ether (2-arachidonyl

glyce-rylether), virodhamine (o-arachidonoylethalamine), and N-arachidonoyldopamine [1,2] Apart from CNRs and their ligands, the endocannabinoid system comprises a specific AEA membrane transporter, a fatty acid amide hydrolase, responsible for AEA degradation to etha-nolamine and arachidonic acid, and an N-acyl-phos-phatidylethanolamines-hydrolyzing phospholipase D,

Keywords

cannabinoid receptor; frog; nonsynonymous

mutations; post-transcriptional modifications;

synonymous mutations

Correspondence

R Pierantoni, Dipartimento di Medicina

Sperimentale, Sez ‘F Bottazzi’, II Universita`

di Napoli, via Costantinopoli 16, 80138

Napoli, Italy

Fax: +39 081 5667536 00

Tel: +39 081 5667617

E-mail: riccardo.pierantoni@unina2.it

Database

The sequences reported in this paper have

been deposited in the GenBank database

under the accession numbers AM113546,

AM260468 and AM260467 for frog cnr1

brain, testis cDNA and genomic DNA,

respectively

(Received 8 January 2007, revised 3 April

2007, accepted 5 April 2007)

doi:10.1111/j.1742-4658.2007.05824.x

The endocannabinoid system is a conserved system involved in the modula-tion of several physiologic processes, from the activity of the central ner-vous system to reproduction Type 1 cannabinoid receptor (CNR1) cDNA was cloned from the brain and testis of the anuran amphibian, the frog Rana esculenta Nucleotide identity ranging from 62.6% to 81.9% is observed among vertebrates The reading frame encoded a protein of 462 amino acids (FCNR1) with all the properties of a membrane G-coupled receptor Alignments of FCNR1 with those of other vertebrates revealed amino acid identity ranging from 61.9% to 88.1%; critical domains for CNR1 functionality were conserved in the frog As nucleotide differences

of cnr1 cDNA were observed in brain and testis, the genomic sequence of the cnr1 gene was also determined in the same tissue preparations Nucleo-tide changes in codons 5, 30, 70, 186, 252 and 408 were observed when cDNA and genomic DNA were compared; the nucleotide differences did not affect the predicted amino acid sequences, except for changes in codons

70 and 408 Interestingly, the predicted RNA folding was strongly affected

by different nucleotide sequences Comparison of cnr1 mRNA sequences available in GenBank with the corresponding genomic sequences revealed that also in human, rat, zebrafish and pufferfish, nucleotide changes between mRNA and genomic sequences occurred Furthermore, amino acid sequences deduced from both mRNA and the genome were compared among vertebrates, and also in pufferfish the nucleotide changes correspon-ded to modifications in the amino acid sequence The present results indi-cate for the first time that changes in nucleotides may occur in cnr1 mRNA maturation and that this phenomenon might not be restricted to the frog

Abbreviations

AEA, arachidonoylethanolamide; CD, cytoplasmic domain; CNR, cannabinoid receptor; ED, extracellular domain; TM, transmembrane domain.

enzymatic machinery that is responsible for the release,

on demand, of AEA from membrane

N-acyl-phospha-tidylethanolamines [3] Currently, two CNR subtypes

have been characterized: type 1 (CNR1) is widely

expressed in the nervous system and in several

per-ipheral tissues, including the pituitary gland and

reproductive tissues [4,5]; type 2 (CNR2) is mainly

expressed in the immune system [6,7] Splice variants

of CNR1 (named CNR1a and CNR1b), with different

pharmacologic effects and expression rates, have been

described in humans [5,8] Furthermore, the presence

of additional CNR subtypes (CNRx) has been

postula-ted in mice [9]

The endocannabinoid system is highly conserved in

evolution Orthologs of the human CNR1 receptor

gene (cnr1) have been cloned and sequenced in

mam-mals [10] Furthermore, cnr1 orthologs have been

cloned and sequenced in fish [11–13], in urodele and

anuran amphibians [14,15], and in birds [16] Reptilian

species have not yet been investigated As well as in

vertebrates, the cnr1 gene has been cloned in an

uro-chordate, the sea squirt Ciona intestinalis [17], and its

high expression has been described in the cerebral

gan-glion, branchial pharynx, heart and testis [18] In

inver-tebrates, the occurrence of endocannabinoid receptor

activity has been reported in sea urchins, molluscs,

anne-lids and cnidarians [19–21] In addition, the

investiga-tion of cnr1 orthologs in the genomes of Drosophila

melanogaster and Caenorhabditis elegans was

unsuc-cessful, and no binding sites for CNR1 synthetic

lig-ands have been found in several insect species [21] In

this respect, the first appearance of the

endocannabi-noid system might be evolutionarily related to that in

deuterostomian organisms [10,21]

CNR1 is a membrane G-protein coupled receptor

with seven transmembrane-spanning regions [22]

The signal transduction pathway elicited by CNR1

comprises inhibition of adenylylcyclase via Gi

pro-tein and consequent activation⁄ inhibition of Ca2+

and K+ channels; activation of mitogen-activated

protein kinase has also been reported [23] Besides

the classical CNR1 and CNR2 receptors, AEA

interacts with K+ and Ca2+ channels, as well as

5-hydroxytryptamine-3 receptor and the type 1

vanil-loid receptor, a ligand-gated and nonselective

cati-onic channel [24] Furthermore, AEA produced after

Ca2+ mobilization has recently been proposed as an

intracellular messenger regulating ion channel activity

by binding the type 1 vanilloid receptor channels on

the cytoplasmic bilayer interface [25] Interestingly,

CNR1 homodimerization and heterodimerization [26]

represent a further amplification of cannabinergic

signalling potency

Therefore, due to the great complexity of the endo-cannabinoid system, the widespread distribution outside the nervous system and the high degree of evolutionary conservation, detailed CNR1 molecular characterization among species may be useful to elucidate the activity of endocannabinoids at multiple levels As a previous report indicates that the endocannabinoid system oper-ates in the brain and testis of the frog Rana esculenta [27], we took advantage of this model to obtain know-ledge of the molecular cloning and characterization of CNR1

We report for the first time the detection of nucleo-tide differences among brain cDNA, testis cDNA and genomic sequences, together with the corresponding amino acid variations We have also investigated whe-ther or not this phenomenon is present in owhe-ther verteb-rate species studied so far

Results

cnr1 molecular cloning and receptor characterization from brain preparations

R esculenta cnr1 partial cDNA was obtained by a combination of RT-PCR and 3¢-RACE (Table 1) The characterized R esculenta brain cnr1 cDNA (fcnr1) is

1586 bp long, and comprises: (a) a 5¢-UTR 12 bp long; (b) a coding region of 1389 bp encoding a protein of

462 amino acids; and (c) a complete 3¢-UTR 169 bp long containing a canonical polyadenylation site at

1522 bp

The fcnr1 nucleotide sequence was compared with those of other vertebrates [mammals (cat, rat, mouse, chimp, monkey and human), amphibians (the anuran Xenopus laevis and the urodele Taricha granulosa), birds (the zebrafinch Taeniopygia guttata) and teleost fish (Fugu rubripes, whose genome contains a cnr1 gene named cnr1a and a paralogous gene named cnr1b, and Danio rerio)] as well as invertebrates (the urochordate sea squirt C intestinalis) cnr1 sequences from other vertebrates and invertebrates containing only a partial coding sequence were not considered in this study Alignments, conducted by lalign and clustalw multiple alignments, revealed a nucleotide identity ran-ging from 62.6% to 81.9% among vertebrates, and 46.5% against C intestinalis (Table 2) A protein of

462 amino acid residues with a predicted molecular mass of 51.89 kDa was deduced from the nucleotide sequence Also, the deduced CNR1 amino acid sequence of R esculenta (FCNR1) was compared with those of known CNR1s, revealing an amino acid iden-tity ranging from 61.9% to 88.1% among vertebrates and of 21.5% in C intestinalis (Table 2) A complete

alignment of known CNR1 proteins is reported in

Fig 1; interestingly, the lowest amino acid identity is

in the N-terminal region of the receptor (amino acids

1–72) (Fig 1)

A rooted phylogenetic tree was constructed using

the phylip drawgram method and exported by

clustalw (Fig 2) On the basis of the estimated

phy-logenetic relationship among the CNR1s in

verte-brates, we confirm a relative divergence between

CNR1 sequences of the anuran and the urodele

amphibians [15]

bioinformaticwas then used for further

characteri-zation of FCNR1 Seven hydrophobic domains, typical

of the G-coupled transmembrane receptor, were

pre-dicted by tmap and tmhmm software (Fig 3) The

four extracellular domains (ED1–4) comprise amino

acid residues 1–109 (ED1), 168–181 (ED2), 248–266

(ED3), and 360–368 (ED4); transmembrane domains

(TM1–7) comprise amino acid residues 110–132 (TM1), 145–167 (TM2), 182–204 (TM3), 225–247 (TM4), 267–289 (TM5), 337–359 (TM6), and 369–391 (TM7); the four cytoplasmic domains (CD1–4) com-prise amino acid residues 133–144 (CD1), 205–224 (CD2), 290–336 (CD3), and 392–462 (CD4)

Although the highest nucleotide and amino acid identity was observed among amphibians, a lower degree of conservation was detected in ED1; in partic-ular, in R esculenta, seven consecutive amino acid resi-dues were completely missing as compared with other amphibian species (Fig 1)

Several putative high-confidence phosphorylation residues (serine, threonine and thyrosine) were predic-ted by netphos 2.0 software (Fig 3)

Critical domains for CNR1 functionality in the other vertebrates were conserved in the frog (Fig 3) Among them were: (a) dual sites for N-linked

Table 1 Primer sequences and PCR programs used for genomic and cDNA fcnr1 amplification.

P1

P2

X laevis

X laevis

CAGTTCTTCCTCTGTTTGGGTGGAAC CCATAAGAGGGCCCCAACAAATG

95 C, 5 min;

94 C for 30 s,

58 C for 45 s,

72 C for 1 min, 35 cycles;

72 C for 7 min

339

P3

P2

Degenerate

R esculenta

GCTTCATGATTCT(GT)A(AC)(CT)CC(AC)AG CCATAAGAGGGCCCCAACAAATG

95 C for 5 min;

94 C for 45 s,

50 C for 45 s,

72 C for 45 s, 5 cycles;

94 C for 45 s,

45 C for 45 s,

72 C for 45 s, 30 cycles;

72 C for 7 s

780

P5

P6

X laevis

R esculenta

AAAACTGGGGTAATGAAGTC AGTAAATGTACCCAGGGTTA

95 C for 5 min;

94 C for 30 s,

50 C for 30 s,

72 C for 45 s, 5 cycles;

94 C for 30 s,

54 C for 30 s,

72 C for 45 s, 35 cycles;

72 C for 7 min

378

P7

P8

R esculenta

Degenerate

ATTGGGGTAACCAGTGTTCT T(GC)GC(AG)ATCTTAAC(AG)GTGCT

95 C for 5 min;

94 C for 30 s,

52 C for 30 s,

72 C for 1 min, 5 cycles;

94 C for 30 s,

48 C for 30 s,

72 C for 1 min, 25 cycles;

72 C for 7 min

528

P9

AP

95 C for 5 min;

94 C for 30 s,

62 C for 1 s,

72 C for 1 min 30 s, 35 cycles;

72 C for 7 min

glycosylation in the N-terminal extracellular domain;

(b) potential sites for protein kinase C phosphorylation

in the first and the third intracellular regions; (c) a

conserved lysine in the third transmembrane domain

(TMD3) whose importance for receptor interaction

with bicyclic but not aminoalkylindole classes of

cann-abinoid agonists has been reported [28,29]; (d) a TQK

motif in the third cytoplasmic loop that is critical in

rat for CNR1 receptor activation of K+ and Ca2+

channels [30]; (e) a leucine and alanine pair in the

C-terminus of the third cytoplasmic loop implicated in

the interaction with Gs in rat [31]; (f) two serine

resi-due within the intracellular tail, corresponding to rat

amino acid residues 426 and 430 involved in receptor

desensitization [32]; (g) a TVK sequence corresponding

to a potential protein kinase C phosphorylation site

within the intracellular tail region; and (h) a TMS

motive in the intracellular tail, corresponding to rat

amino acid residues 460–463, required for

WIN55212-2-mediated receptor internalization in

AtT20-trans-fected cells [33]

fcnr1 molecular cloning from testis preparations

A fragment of 1384 bp (codons 1–447) was cloned by RT-PCR from frog testis Alignment between R escu-lenta brain and testis cDNA revealed two nucleotide differences in codons 186 (GGG in testis and GGA in brain) and 252 (CTC in testis and CTA in brain) Such modifications were constantly observed from the sequences of three different clones isolated from differ-ent cDNA preparations and did not correspond to amino acid differences

fcnr1 genomic DNA sequence analysis

To assess the possibility of post-transcriptional modi-fications, we cloned the whole R esculenta coding region of cnr1 from genomic DNA preparations obtained from the same homogenates used to pre-pare cDNA Similar results were obtained from genomic DNA sequences from testis, brain and muscle clustalw alignments of brain cDNA, testis

Table 2 Nucleotide and amino acid identity (%) between the frog Rana esculenta CNR1 receptor and other CNR1 and CNR2 receptors Nuc-leotide identity is referred to the coding sequences Accession numbers in the NCBI GenBank for cnr1 nucNuc-leotide sequences: Ciona intesti-nalis, AB087259; Fugu rubripes cnr1a, X94401; Fugu rubripes cnr1b, X94402; Danio rerio, AY148349; Taricha granulosa, AF181894; Xenopus laevis, AY098532; Taeniopygia guttata, AF255388; Mus musculus, AF153345; Rattus norvegicus, U40395; Felis catus, U94342; Macaca mulatta, AF286025; Pan troglodytes, NM_001013017; Homo sapiens, NM_016083; Homo sapiens cnr1a, NM_033181; Homo sapiens cnr1b, AY766182 Accession numbers in the NCBI GenBank for cnr2 nucleotide sequences: Danio rerio, NM_212964; Mus musculus, NM_009924; Rattus norvegicus, NM_020543; Homo sapiens, NM_001841.

CNRs

% Nucleotide identity

Coding length (nucleotides)

% Amino acid identity

Amino acid residues Cnr1

Cnr2

cDNA and genomic sequence revealed several

nuc-leotide differences (Fig 4) Brain cDNA differed

from the genomic cnr1 sequence in codons 5, 30, 70,

186, 252, and 408 Testis cDNA differed from the

genomic cnr1 coding sequence in codons 5, 30, 70,

and 408 Owing to genomic code degeneration, these

nucleotide differences did not change the amino acid

sequence except for those concerning codon 70 and

codon 408 In fact, at the genomic level, TCA70 and

AAA408 encoded serine (S70) and lysine (K408),

respectively; in the cDNA, GCA70 and GAA408

encoded alanine (A70) and glutamic acid (E408),

respectively A70 is located in the first extracellular

loop, just in the conserved N-linked glycosylation

domain; E408 is located in the cytoplasmic tail, in a

region suggested to be sensitive for Gicoupling in

rat [23] To assess whether or not similar nucleotide

differences among cnr1 cDNA and the corresponding

genomic sequences exist in vertebrates, we blasted

the mRNA sequences deposited in GenBank against

the corresponding genome database Amino acid sequences deduced from mRNA sequences were then compared to those deduced from the corresponding genomic sequences The results of our search are summarized in Table 3 For rat, human, zebrafish and pufferfish cnr1b, differences between mRNA and genomic sequences are reported In particular, in

F rubripes cnr1b, such nucleotide differences also corresponded to amino acid differences

Northern and Southern blot analysis Northern blot analysis of R esculenta brain and testis mRNA was carried out using an antisense RNA probe

of 780 bp

A signal of 2.2 kb was observed in brain and testis (Fig 5A) Southern blot analysis revealed a single sig-nal of 3 kb in frog genomic DNA, previously digested

by EcoRI, indicating that fcnr1 is a single-copy gene (Fig 5B)

Fig 1 Alignments of complete CNR1 amino

acid sequences Completely conserved

amino acid residues are in black boxes;

identical amino acid residues are in light

gray boxes; similar amino acid residues are

in medium gray boxes; different amino acid

residues are in white boxes Similarity ⁄

dif-ferences have been highlighted with the

BOX - SHADE alignments graphic program.

RNA folding analysis

A difference between brain and testis mRNA

secon-dary structure emerged (Fig 6A,B, asterisks); several

differences in the secondary structure predicted from

brain⁄ testis mRNA and the mRNA sequence deduced

from genomic DNA were also observed (Fig 6C,

arrows)

Discussion

In this article, we report the molecular cloning of cnr1

from R esculenta brain and testis fcnr1 and FCNR1

have high nucleotide and amino acid identity (ranging from 62.6% to 81.9% and from 61.9% to 88.1%, respectively) as compared to those of other vertebrates Several critical domains for CNR1 functionality are present in frog, suggesting an evolutionarily conserved activity Furthermore, analysis of cnr1 gene organiza-tion suggests that the cnr1 coding region is contiguous and not interrupted by intronic sequences as reported for other vertebrates [5,8,34] In fact, splice donor– acceptor sites detected in mouse, rat and human, responsible for cnr1a and cnr1b splice forms, are not conserved in frog Interestingly, comparison between genomic DNA and cDNA, both obtained from frog

Fig 1 (Continued).

brain and testis, suggests the existence of nucleotide changes in cDNA sequences Four single-nucleotide polymorphisms responsible for the modulation of stria-tal response to happy faces have been reported in the human cnr1 gene [35,36] In the present study, we think that the possibility of polymorphic sites may be excluded, in that DNA and RNA preparations were derived from the same tissue preparations collected from five animals per month at least, and we always confirmed our results Furthermore, sequences obtained from brain and testis genomic DNA were identical In addition, to avoid any sequence deduction from 3¢-overlapping ends, sequencing was conducted

on both strands from three separate clones no more than 800 bp long Interestingly, only alterations in codons 70 and 408 are effective in changing amino acid residues A70 and E408 are located in the first extracellular domain, just in the conserved N-linked glycosylation domain, and in the cytoplasmic tail, a region suggested to be sensitive for Gi coupling in rat

Fig 3 Frog CNR1 receptor characterization Extracellular domains are in Courier New; transmembrane domains are in bold characters; intra-cellular domains are in italics *Putative high-confidence phosphorylation sites; dual sites for N-linked glycosylation in the N-terminal extra-cellular domain are underlined ^The conserved lysine in the third transmembrane domain (TMD3) Leucine and alanine pair in the C-terminus of the third cytoplasmic loop implicated in the interaction with Gsin rat; the light gray box indicates the TQK motif in the third cytoplasmic loop that is critical in rat for CB1 receptor activation of K + and Ca 2+ channels; the medium gray box indicates two serine resi-dues within the intracellular tail that in rat are phosphorylated and involved in receptor desensitization; the dark gray box indicates a TVK sequence corresponding to a potential protein kinase C phosphorylation site within the intracellular tail region; the black box indicates a TMS motive in the intracellular tail that is required in rat for WIN55212-2-mediated receptor internalization in AtT20-transfected cells; the white boxes indicate possible editing sites, and different amino acid residues predicted from brain cDNA and genomic DNA are in bold italic characters.

Fig 2 Phylogenetic analysis of the known vertebrate CNR1

recep-tor A rooted phylogenetic tree was constructed using the PHYLIP’S

DRAWGRAM method and exported by CLUSTALW Branch lengths are

proportional to the estimated evolutionary distance among the

receptors.

[23]; in this respect, a role in the modulation of CNR1

activity is not excluded Finally, Southern and

northern blot analysis demonstrate that cnr1 is a

single-copy gene, and therefore the corresponding mes-senger is detected in both brain and testis

To verify whether or not cnr1 differences between ge-nomic and cDNA sequences occur in other vertebrates,

we blasted all the nucleotide sequences deposited in NCBI GenBank with the corresponding genomic avail-able sequences; in addition, we compared all the amino acid sequences deduced from both genomic DNA and mRNA It is worth noting that similar nucleotide chan-ges occur in other species and are quite scattered among vertebrates However, as in the frog, most changes do not influence the amino acid composition Only in the

F rubripes cnr1bgene do nucleotide changes, in codons

241 and 463, change the amino acid composition pre-dicted by the analysis of different genomic DNA sequences

In both mammals and invertebrates, RNA editing

is an elaborate and precise form of post-transcrip-tional RNA processing, powering genetic diversi-fication [37] In mammals, there are two main classes

of editing enzymes able to deaminate encoded nucleotides: the former generates I (inosine) from A (adenosine), and the latter generates U (uridine) from

C (cytidine) [38,39] In the first case, genomically encoded A is read as G in RNA-cDNA sequences Currently, the literature concerning mRNA editing is limited to a relatively few examples In particular, all currently known A-to-I edited transcripts of both mammals and invertebrates encode membrane pro-teins in nervous tissue These propro-teins function as voltage-gated or ligand-gated ion channels or as

G protein-coupled receptors [e.g serotonin (5-hy-droxytryptamine 2c) receptor, non-N-methyl-d-aspar-tate glutamate receptor channels in mammals, and squid K+channels in invertebrates] [40] Interestingly,

if the editing process occurs in frog, a GAA408 codon corresponds to the genomic AAA408 in the cDNA, and this might be considered as an A-to-I editing example With respect to the significance of nucleotide changes in cDNA that do not change the amino acid composition predicted by the genomic sequence, we have no explanation at present Synony-mous mutations in the human dopamine receptor D2

Fig 4 Differences among fcnr1 sequences obtained from genomic DNA and cDNA isolated from brain and testis Alignment of frog cnr1 nucleotide sequences from frog brain genome and brain and testis cDNA Deduced amino acid sequences are in italics Codons with nucleo-tide differences are underlined; bold characters indicate nucleonucleo-tide and amino acid differences.

Table 3 cnr1 mRNA versus cnr1 genomic DNA in vertebrates:

nuc-leotide (nt) and deduced amino acid sequence difference

compar-ison Accession numbers in the NCBI GeneBank for cnr1 mRNA

sequences are the same as in Table 2; putative Fugu rubripes

cnr1a and cnr1b mRNA sequences were deduced from the cloned

gene sequences [11] Accession numbers in the NCBI

Gene-Bank for cnr1 genomic sequences: Fugu rubripes cnr1a,

CAAB01001440.1; Fugu rubripes cnr1b, CAAB01001484.1; Danio

rerio, NW_634056.1; Mus musculus, NT_109315.2; Rattus

norvegi-cus, NW_047711.2; Felis catus, NM_001009331.1; Pan troglodytes,

NW_107960.1; Homo sapiens, NT_086697.1.

Species

Genome mRNA (nucleotides)

Genome cDNA (amino acids)

brain testis

2.2 -

KB

-3.0 KB

18S -

Fig 5 Northern blot and Southern blot analysis (A) A band of

2200 bp is observed in both brain (1) and testis (2) mRNA by

nor-thern blot experiments (B) A single band of 3000 bp is observed

from high-mass genomic DNA previously digested with EcoRI,

indi-cating that fcnr1 is a single-copy gene.

affect mRNA stability and synthesis of the receptor

[41] Accordingly, in the present study, synonymous

and nonsynonymous mutations weakly alter the

puta-tive mRNA folding between brain and testis mRNA;

by contrast, the secondary structure of the mRNA

predicted from the genomic sequence is substantially

different from the secondary structure of brain and

testis mRNA In this respect, we speculate that the

nucleotide changes observed in this study may affect

RNA folding and therefore its stability and turnover

In conclusion, apart from molecular cloning of cnr1

in R esculenta, this is the first report showing that

changes in nucleotides occur during mRNA

matur-ation Furthermore, we find that this phenomenon is

not restricted to the frog Whether or not editing

pro-cesses may generate different cnr1 mRNA molecules

(with different activity⁄ stability) in different tissues,

leading to pharmacologic applications, should be

fur-ther investigated

Experimental procedures

Animals and tissue collection

Five male frogs of R esculenta were collected monthly

from September until July in the neighborhood of

Naples (Italy) The animals were killed under anesthesia

with MS222 (Sigma-Aldrich Corp., St Louis, MO)

Brain, testis (for genomic and cDNA studies) and

mus-cle were removed and stored appropriately at ) 80 C

until used This research was approved by the Italian

Ministry of University and Scientific and Technological

Research

Total RNA preparation

A pool of total RNA was extracted from R esculenta tis-sues (n¼ 5 per month) using TRIZOL Reagent (Invitrogen Life Technologies, Paisley, UK), following the manufac-turer’s instructions Total RNA was treated for 30 min

at 37C with DNaseI (10 U per sample) (Amersham Pharmacia Biotech, Chalfont St Giles, UK) to avoid any contamination of genomic DNA; total RNA purity and integrity were determined by spectrophotometer analyses at

260⁄ 280 nm and by electrophoresis

Isolation of R esculenta complete cnr1 coding sequence

Pools of total RNA were reverse transcribed to prepare cDNA The reverse transcription was carried out using

5 lg of total RNA, 0.5 lg of anchor oligonucleotide (AP),

10 mm dNTP, 0.01 m dithiothreitol, 1· first-strand buffer (Invitrogen Life Technologies), 40 U of RNase Out (Invi-trogen Life Technologies), and 200 U of SuperScript-III RNaseH– reverse transcriptase (Invitrogen Life Techno-logies), in a final volume of 20 lL, following the manufac-turer’s instructions As negative control, total RNA not treated with reverse transcriptase was used

Complementary DNA was used for PCR analysis A frog cDNA fragment of 339 bp corresponding to trans-membrane segments 4 and 7 was obtained using primers designed from X laevis cnr1 cDNA (P1 and P2, Table 1 for details) Afterwards, rat, mouse, African clawed frog and zebrafish cnr1 complete nucleotide sequences were aligned by clustalw, and degenerate upper and reverse primers were selected in highly conserved regions To extend the 5¢ ⁄ 3¢ sequence, combinations of

degener-Fig 6 mRNA folding analysis Secondary

structure of the overlapping fragments of

1354 bp from brain mRNA (A), testis mRNA

(B) and the mRNA sequence deduced from

genomic DNA (C) Hatched circles mark the

regions with the main differences between

the cloned mRNA and the mRNA sequence

deduced from genomic DNA Asterisks

indi-cate the difference between brain and testis

mRNA; arrows indicate the main differences

in mRNA deduced from genomic DNA Bold

circles mark the magnification of the

secon-dary structure of brain and testis mRNA.

Structures have been selected on the basis

of minimal dG (free energy) values.

ate⁄ specific primers were used Primer sequences and

PCR program details are summarized in Table 1 All

PCR analyses were conducted in an Applied Biosystem

Thermocycler apparatus using 1 lL of diluted cDNA or

100 ng of genomic DNA and the high-fidelity TaKaRa

Ex Taq (Cambrex Bio Science, Milan, Italy) Amplification

products were subcloned in pGEM-T Easy Vector (Promega

Corporation, Madison, WI) DH5a high-efficiency

compet-ent cells were transformed, and recombinant colonies were

identified by blue⁄ white color screening Plasmidic DNA

was extracted by NucleoBond Plasmid extraction kit

(Macherey-Nagel, Du¨ren, Germany), and insert size was

controlled by restriction analysis with EcoRI (Fermentas

GmbH, St Leon-Rot, Germany) DNA was then

se-quenced on both strands by Primm Sequence Service

(Primm srl, Naples, Italy) Finally, frog cnr1 mRNA and

genomic nucleotide sequences were obtained by

compar-ing overlappcompar-ing fragments, and amino acid sequences

were deduced

Sequence analysis

Nested amplification products, obtained independently

from separate amplification reactions, were sequenced in

both forward and reverse strands, and the cnr1 mRNA

sequence was deduced by comparing overlapping fragments

of the two complementary strands

On the basis of the nucleotide sequence, the R esculenta

CNR1 amino acid sequence was deduced In order to

esta-blish the degree of CNR1 identity among vertebrates, both

the nucleotide coding sequence and the amino acid

sequence of R esculenta CNR1 were aligned with other

known CNR1⁄ CNR2 complete coding sequences and

amino acid sequences available in the NCBI GenBank by

alignand clustalw multiple alignments

After alignments of vertebrate CNR1 amino acid

sequences, a phylogenetic tree was constructed using phylip’s

drawgramand exported from clustalw

Finally, putative transmembrane domains, N-linked

gly-cosylation sites and phosphorylation sites were predicted by

using, respectively, tmap, tmhmm, netnglyc 1.0 Server,

and netphos 2.0 Server, available at SDSC Biology

Work-bench (http://workWork-bench.sdsc.edu/) and at ExPASy

Proteo-mics Server (http://au.expasy.org/)

cnr1 ORF prediction from genomic sequences

of several vertebrate

To assess the presence of nucleotide differences between

cDNA and the corresponding genomic sequences in

vertebrates, cnr1 cDNA sequences deposited in NCBI

GenBank were blasted against the corresponding

geno-mic database (see Table 3 for details) Amino acid

sequences were deduced from the corresponding genomic

cnr1 sequences and compared to those predicted from cDNA

Northern blot Ten micrograms per lane of total brain and testis RNA pre-viously denatured with glyoxal and dimethylsulfoxide was electrophoresed on 1.4% agarose gel and blotted on nylon membranes (Nytran; Amersham Pharmacia Biotech) An antisense RNA probe complementary to the cloned 780 bp fragment of fcnr1 mRNA was produced using the nonradio-active, digoxigenin (DIG)-based system DIG RNA Label-ling Kit (SP6⁄ T7) (Roche, Mannhein, Germany), following the manufacturer’s instruction In brief, 6 lg of pGEM-T-fcb1 recombinant plasmid was linearized by digestion with Pst1 (Fermentas GmbH), and transcription was carried out

in vitrousing T7 RNA polymerase and a mixture of dNTP and DIG-11-UTP

Blots were prehybridized and probed at 65C in Church’s buffer (0.5 m NaCl⁄ Pi, pH 7.4, 7% SDS, 0.5 mm EDTA, and 100 mgÆmL)1 sonicated salmon sperm); 100 ngÆmL)1 labeled probe was then added in hybridization buffer The membrane was washed twice at room temperature for 5 min

in low-stringency buffer (2· NaCl ⁄ Cit, 0.1% SDS), and then twice at 65C in high-stringency buffer (0.2 · NaCl ⁄ Cit, 0.1% SDS) The chemiluminescent protocol for the detection of DIG-labeled probes suggested by Roche was then used After incubation with the chemiluminescent alka-line phosphatase substrate CSPD (Roche), the filter was exposed for a suitable time to Hyperfilm Kodak (Rochester, NY) autoradiographic film

Genomic DNA extraction and Southern blot analysis

High molecular mass DNA was extracted and purified by standard techniques [42] from frog tissues (n¼ 5 per month) Genomic DNA extracted from muscle (10 lg) was digested with EcoRI, BamHI or HindIII (Fermentas GmbH) and analyzed by electrophoresis and Southern blot using a 780 bp frog fcnr1 DIG-labeled probe The probe was labeled by random priming, and hybridization was conducted at 50C using the DIG-High Prime DNA Labelling and Detection Starter Kit II (Roche), following the manufacturer’s instruction

mRNA folding analysis Prediction of the secondary structure of RNA and DNA was conducted by using mfold software, available at http://www.bioinfo.rpi.edu/applications/mfold/rna/form1.cgi [43,44] This analysis was restricted to the overlapping frag-ments of 1354 bp from testis mRNA, brain mRNA and genomic DNA