Proc. Natl. Acad. Sci. USAVol. 93, pp. 12344-12348, October 1996Cell Biology

Identification of a new calcitonin gene in the salmonOncorhynchus gorbuscha

(salmon calcitonin gene/salmon calcitonin IV/calcitonin gene-related peptide IV)

HENK JANSZ*, KELLY MARTIALt, JOHANNA ZANDBERG*, GEIRARD MILHAUDt, A. A. BENSON§, ANNICK JULIENNEt,M. S. MOUKHTARt, AND MICH#ELE CRESSENTT¶*Laboratory for Physiological Chemistry University of Utrecht, 3584 CG Utrecht, The Netherlands; tU.349 Institut National de la Sante et de la Recherche, 6 rueGuy Patin, centre Viggo Petersen, 75010 Paris, France; tService de Biophysique, H6pital Saint Antoine, 184 Rue du Faubourg Saint Antoine, 75012 Paris, France;and §Scripps Institution of Oceanography, La Jolla, CA 92093-0202

Communicated by Andrew A. Benson, University of California at San Diego, La Jolla, CA, July 3, 1996 (received for review February 18, 1995)

ABSTRACT Three isoforms of calcitonin (CT) exist insalmonids. Isohormones I and II are expressed in the pinksalmon Oncorhynchus gorbuscha. We report here the existencein this species ofa CT gene and of its transcripts, which encodefor a fourth isohormone, the salmon CT (sCT) IV. This newCT gene was identified by PCR from genomic DNA and bysequencing the amplified DNA. The expression of this CT genewas established in ultimobranchial body and brain, by reversetranscription-PCR, hybridization and sequencing. The sCT IVgene, like the sCT I gene, is a complex transcription unit,containing exons encoding for a CT as a calcitonin gene-related peptide (CGRP) molecule. The predicted peptide, sCTIV, has a greater homology with the eel CT and the sCT II thanwith the sCT I. Alignment of the sCT IV with other fish andchicken CT showed amino acid modifications in similarpositions as those found during evolution. The predictedsalmon CGRP IV peptide is highly homologous to the knownCGRP molecules in other species, confirming the high con-servation of the molecule during evolution. This identificationofa new salmon CT gene is interesting both for the therapeuticpotential represented by the new molecules encoded by thisgene and for phylogenetic studies.

The calcitonin (CT) gene is a complex transcription unitencoding both the CT peptide and the calcitonin gene-relatedpeptide (CGRP), which are produced by alternative splicing ofthe primary transcript (1). Identical organization was estab-lished in rat, human, chicken, and salmon species (1-5).Existence of a second CT gene was then demonstrated in ratand human (6, 7). However, in these two studied species, onlythe gene I encodes for a CT molecule, the gene II beingunproductive for the CT peptide.

Study of the CT genes in a nonmammalian vertebrate, thesalmonid species, is particularly interesting, as in this species,three isohormones of CT have been identified (8), and,paradoxically, two of them are the most biologically activemolecules in mammals (9). So far, only the gene encoding forthe isohormone I has been characterized. The nucleotidesequence of the salmon CT (sCT) I cDNA was primarilyestablished (10), and, recently, genomic DNA analysis re-vealed that a CGRP molecule can also be encoded by the samegene (5), as in mammalian vertebrates. Existence in salmon ofsuch a complex transcription unit coding for the sCT I allowedus to speculate that similar genes coding for the other sCTisohormones might exist, and so other salmon CGRP (sCGRP)molecules could also occur in that species.

In view of identifying these CT gene sequences, we pursuedour studies on salmon genomic DNA. We were encouraged inthis search by the fact that during identification of the sCT

I/sCGRP I gene in salmon DNA (5), evidence was obtainedfor the existence of other CT/CGRP genes. These researchesled us to identify a new CT gene in salmon. This gene bothencodes for a CT and a CGRP molecule, whose sequences hadnot yet been elucidated. We further looked for expression ofthese two molecules in salmon tissues. Comparison of theirdeduced amino acid sequences with those of other vertebratesprovides a new insight in evolution of the CT gene family andsupports biological interest of these new CT and CGRPmolecules.

MATERIALS AND METHODSExperimental Procedures. Genomic DNA of salmon testes

(Oncorhynchus keta) was purchased from Pharmacia (productno. 27-4564, lot no. BJ 4564101).

Fresh tissues were collected from male and female pinksalmon (Oncorhynchus gorbuscha) during an expedition toAlert Bay at the north end of Vancouver Island, BritishColumbia. After netting, the fish were immediately killed by asharp blow to the head and ultimobranchial bodies (UBs) andbrains were removed, frozen in liquid nitrogen, and stored at-800C.RNA Extraction from Tissues. Total RNA was extracted

from UBs and brain tissues by a single-step extraction withguanidinium thiocyanate and phenol/chloroform (11). Theywere quantified using an OD of 260 nm and stored at -80°C.PCR Amplification of Genomic DNA and DNA Fragments

Analysis. In previous study (5), during identification of the sCTI/sCGRP I gene using direct and inverse PCR amplificationand universal primers, evidence was obtained for the existenceof a second sCT gene (called sCT IV gene) in salmon DNA.Based on a provisional sequence of this gene, we were able toconstruct specific primers for PCR amplification of regionsencoding sCT IV and sCGRP IV peptides, respectively (Table1). PCR was performed essentially as described by Saiki et al.(12). The thermal profile was as follows: denaturation for 1 minat 94°C, annealing for 2 min at 55°C, and extension for 30 secat 72°C, for 40 cycles. After electrophoresis of the amplifiedproducts, DNA fragments of the expected length was extractedfrom 3% Nusieve agarose gel (FMC), cleaved by BamHI, andcloned in M13 mp8 as described (5). DNAs from at least 10clear plaques were sequenced by the dideoxy chain terminationmethod (13) using the T7 Sequencing kit (Pharmacia) and[35S]-ATP (Amersham).

Abbreviations: CT, calcitonin; CGRP, calcitonin gene-related peptide;sCT, salmon CT; sCGRP, salmon CGRP; UB, ultimobranchial body;RT-PCR, reverse transcription PCR; oligo, oligonucleotide.Data deposition: The sequences reported in this paper have beendeposited in the GenBank data base [accession nos. U71286 (for CTIV) and U71287 (for CGRP IV)].1To whom reprint requests should be addressed.

12344

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement" inaccordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 14

, 202

2

Proc. Natl. Acad. Sci. USA 93 (1996) 12345

Table 1. Primers used for PCR amplification of genomic DNA

Oligos Sequence 5' -> 3'

CT IV 9145 AGCATGGAAAGGCCCATTACT IV 9146 TTGCATAGCGTTCGCTCTCGCGRP IV 9147 TCCTGAAGCAGCATTACAACGRP IV 9148 TGGAATGTTCACATCTGTG

To facilitate cloning, each primer was equiped with a linker con-sisting of partially overlapping PstI/BamHI sites (CTGCAGGATCC).

Subsequently, to demonstrate that the two amplified CT andCGRP molecules belong to the same gene, a PCR amplifica-tion using the primer set 9145/9148 (Table 1) was performedas described (5). The amplified DNA fragments were thencloned and sequenced from both ends as described above.

Reverse Transcription-PCR (RT-PCR) Amplification ofUBand Brain mRNA. UB and brain mRNAs were transcribed tocDNAs with Moloney murine leukemia virus reverse trans-criptase (BRL) and a synthetic primer, the poly(dT)17 boxprimer [oligonucleotide (oligo) 42; Table 2]. Then, the first-strand cDNAs was amplified using Taq polymerase (Eurobio,Paris), a 3' primer identical to the end of the poly(dT) boxprimer (oligo 41; Table 2) and a 5' primer specific for exon 4of the sCT IV genomic sequence chosen upstream of the CTcoding sequence (oligo S'4; Table 2). A second PCR ampli-fication was performed using the same 5' primer and a 3'primer specific for exon 4 of the sCT IV genomic sequencesituated downstream to the CT coding sequence (oligo S4;Table 2). A similar protocol was used to establish expressionof sCGRP IV molecules in UB and brain tissues. In that case,cDNAs were amplified using primarily a 5' primer specific forthe first bases of the sCGRP IV genomic sequence (oligoCG'4; Table 2) and the 3' primer, oligo 41 (Table 2). Thesecond PCR was performed with the previous 5' primer (oligoCG'4) and a 3' primer specific for the end of the exon 5genomic sequence (oligo CG4; Table 2).Adequate controls were included, and they consisted of

samples with RNase treatment or without the reverse trans-criptase step. The amplification conditions were as follows:denaturation for 1 min at 94°C, annealing for 1 min at 55°C,and further extension for 1 min at 72°C, for 30 cycles in athermal cycler. Purified PCR primers were purchased fromGenosys (United Kingdom).

Southern Analysis. The above PCR amplification productswere run on either a 1.3% or a 2% agarose gel. Size estimationswere deduced from DNA molecular mass marker (BoehringerMannheim). Blotting onto a Gene-screen membrane (NEN) ofthe PCR products was achieved as described (14). The Gene-screen membrane with the transferred PCR products obtained

Table 2. Nucleotide sequences of the different oligos used asprimers for RT-PCR amplification of UB and brainmRNA and as probes

Oligos* Sequence 5' -+ 3't

Oligo 42 GACITCGAGTCGACATCGAT17Oligo 41 GACTCGAGTCGACAT CGOligo S'4 CATGGAAAGGCCCATTAOligo S4 TTAGATGCTGTCAAATGTCOligo SR TGCTCCAACCTCAGCACCTGTOligo CG'4 GAGCCTGTAACACAGCCOligo CG4 TCACATCTGTGTGTTTCTC

*Oligos 42, 41, S4, and CG4 were antisense oligos, while the otherswere sense oligos.tOligo 42 contained 17 d(T) residues, which matched with the poly(A)sequence of mRNA and an additive sequence, the box sequence.Oligo 41 was homologous to the box sequence of oligo 42. Oligos S'4and S4 were a set of primers specific for exon 4 of sCTIV gene, whilethe set of oligos CG'4 and CG4 was specific for exon 5. Oligo SR, usedas a probe, was specific for exon 4 of sCT I mRNA sequence andcrosshybridized with exon 4 of sCT IV sequence.

after RT-PCR amplification with specific primers of sCT IVwas hybridized with a radiolabeled oligo probe correspondingto the first 21 nt of the sCT I sequence, which cross-hybridizedwith the corresponding region of the sCT IV sequence (oligoSR; Table 2).

Oligo probe was end-labeled with [32P]-ATP (NEN) and T4kinase (BRL). Hybridizations of the Gene-screen membraneswere performed as described (14).

Sequence Analysis of the RT-PCR Fragments. PCR prod-ucts obtained after RT-PCR amplification of UB mRNA withspecific primers of sCT IV (oligo S'4 and oligo 41) weresubcloned into a plasmid vector using the TA cloning system(Invitrogen) according to the manufacturer's instructions.Plasmid DNAs were sequenced by the dideoxy nucleotidetermination method (13) using the dsDNA cycle sequencingsystem as described by the supplier (GIBCO/BRL) and 33P-labeled forward primer of the vector as sequencing primer. Adirect sequence analysis of the PCR product was also per-formed to circumvent Taq polymerase error. For that, theDNA fragment obtained after the second round of PCR(primers set S'4 and S4, Table 2) was preliminarily fractionatedon agarose, extracted using the QIAEX kit (Qiagen, Chats-worth, CA), and sequenced.PCR products obtained after amplification of brain cDNA

with specific sCGRP IV primers were subjected to the sameprotocols as described above, except that only DNA fragmentscorresponding to the second round of PCR (primer set: G'4,G4) were used in that case.Alignment of the Nucleotide Sequences of sCT IV and

sCGRP IV with Those of sCT I and sCGRP I. Homologystudies between the sCT IV and sCT I nucleotide sequenceswere carried out according to the method of Kanehisa (15)using standard parameters. Similar studies were performed forthe sCGRP IV and CGRP I molecules.Comparison of the sCT IV and sCGRP IV Amino Acid

Sequences with Those of Other Vertebrates. The amino acidsequences translated from the nucleotide sequences werecompared with those of other vertebrates using the alignmentprogram of protein sequence (15).

Phylogenetic Study. A phylogenetic analysis was performedwith all the known CT amino acid sequences.

RESULTSIdentification of a New CT Gene by PCR Amplification of

Genomic DNA and Sequence Analysis. PCR amplificationusing sCT IV specific primers 9145/9146 and salmon genomicDNA as template produced a DNA fragment whose deducedamino acid sequence corresponded to a typical CT peptide(Fig. 1A). The 32-aa CT peptide was located between thecleavage site, Lys-Arg, and the cleavage-amidation sites, Gly-Lys-Lys-Arg, as in all known CT precursors. A 5-aa differencewas observed between this peptide and the sCT I peptide.These amino acid modifications were situated in positions 15,22, 26, 27, and 29. In addition, some rare silent modificationsin nucleotides were also noted in the sCT coding sequence(Fig. 1A).With the sCGRP specific primers 9147 and 9148, a DNA

fragment encoding a CGRP molecule was obtained (Fig. 1B).The 37-aa CGRP peptide was between a Lys-Arg cleavage siteand a Gly-Arg-Arg-Arg-Arg cleavage-amidation site as inother CGRP precursors. Many nucleotide variations wereobserved between the sCGRP IV and the sCGRP I sequences,all of which were silent at the amino acid level, except for theone in position 17. The 3' coding sequence of exon 5, situateddownstream to the CGRP sequence, was 4 aa smaller than thatof the sCGRP I (Fig. 1B).PCR amplification, using the primer set 9145/9148 and

salmon genomic DNA as template, produced a DNA fragmentof - 1100 bp. Sequence analysis of this DNA fragment allowed

Cell Biology: Jansz et al.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 14

, 202

2

Proc. Natl. Acad. Sci. USA 93 (1996)

(A)3 0

Met Glu Arg Pro Ile Thr Las_Ara Cys Ser Asn Leu Ser Thr CysATG GAA AGG CCC ATT ACC AAG CGC TGC TCT AAC CTC AGC ACC TGCC T A G T T C TLeu Asp Met Ser

60 90Val Leu Gly Lys Leu Ser Gln Asp Leu His Lys Leu Gln Thr PheGTG CTG GGC AAA CTG TCC CAG GAC CTG CAC AAA TTA CAA ACG TTT

A G T G G ACGlu Tyr

120Pro Arg Thr Asp Val Gly Ala Gly Thr Pro Glv r_ Lvs A SerCCC CGC ACG GAC GTG GGC GCG GGC ACG CCT GGC AAG AAA CGC AGC

C A AC A AGTAsn Thr Ser

150Ala Pro Glu Ser Glu Arg Tyr Ala Ser Tyr Gly Lys Thr PheGCG CCC GAG AGC GAA CGC TAT GCA AGC TAC GGG AAG ACA TTTTTG C A C A G C T ALeu Asn Asp Ser Tyr

180Asp Ser Ile StopGAC AGC ATC TAA

T G A GGly

()30

Ile Thr Thr Gln ys Ara Ala Cys Asn Thr Ala Thr Cys Val ThrATT ACA ACA CAG AAG CGA GCC TGT AAC ACA GCC ACC TGC GTG ACTG G T G C T T A T C CVal Ala

60 90His Arg Leu Ala Asp Phe Leu Ser Arg Ser Gly Gly Met Gly AsnCAC CGC CTA GCA GAT TTT CTG AGC CGG TCG GGG GGC ATG GGC AAC

G T C C A A A A A AAsn

120Ser Asn Phe Val Pro Thr Asn Val Gly Ala Lys Ala Phe Gly_A&aAGT AAC TTT GTC CCC ACT AAC GTA GGA GCT AAG GCT TTT GGC CGG

C C G C C G C A A

150=Azgg...&g~.Asn Thr Gln Met StopCGG AGG AGA AAC ACA CAG ATG TGAAA G GC CC A AC GCA CCT CTG TA

Asp Ser Pro Lys Thr Ala Pro Leu Stop

FIG. 1. Nucleotide sequences and deduced amino acid sequencesof exon 4 (A) and exon 5 (B) of the sCT IV gene. The deduced aminoacid sequences are indicated above the nucleotide sequences. The32-aa sCT IV peptide is between the Lys-Arg cleavage site and theGly-Lys-Lys-Arg cleavage-amidation site (underlined). The 37-aminoacid sCGRP IV peptide is between the Lys Arg cleavage site and theGly-Arg-Arg-Arg-Arg cleavage-amidation site (underlined). Belowthe sCT IV and the sCGRP IV nucleotide sequences differences innucleotides and corresponding amino acids of sCT I and sCGRP 1 (5,10) compared with those of sCT IV and sCGRP IV are listed. Adifference of 5 aa is observed between the two sCT I and sCT IVpeptides, in positions 15, 22, 26, 27, and 29. The two sCGRP sequencesshow one amino acid modification in position 17. Numbers above thepeptides sequences refer to nucleotide position. Nucleotide sequences1-16 (A) and 149-159 (B) are provisional, since they were derivedfrom primers 9145 and 9148 (Table 1), respectively.

us to obtain nucleotide sequences that exactly matched withthe sCT and sCGRP sequences presented in Fig. 1 A and B,respectively. This result permitted the conclusion that the twosequences belong to one gene. This conclusion was supportedby the fact that the distance between these two sequences insalmon DNA (-1100 bp) was similar to the distance in humanDNA (1300 bp). In analogy with the known CT/CGRP genes,the available evidence suggested that the sequences in A andB in Fig. 1 represented exon 4, encoding sCT IV, and exon 5,encoding sCGRP IV, respectively, of this gene.

Identification in UB and Brain of sCT IV and sCGRP IVmRNA by RT-PCR. After PCR amplification of cDNA, ob-tained by reverse transcription of UB and brain RNAs with a5' primer specific for sCT IV and a 3' oligo d(T)17 box primer,two DNA fragments were obtained for each tissue (Fig. 24).Hybridization to the SR oligo probe allowed to identify theDNA band of the lower size, "600 bp, as the specific amplifiedPCR product (Fig. 2B).

1 2 3 4 5 6 7 8

A B

FIG. 2. (A) Southern blot analysis of PCR products obtained afteramplification of UB and brain cDNAs with a 5' primer specific of sCTIV (oligo S'4; Table 2) and a 3' primer corresponding to the end of thepoly(dT) box primer (oligo 41; Table 2). Lanes 2-5, controls; lane 6,size marker; and lanes 7 and 8, DNA fragments obtained afteramplification of UB and brain cDNAs, respectively. (B) Autoradiog-raphy of the Southerns blot hybridized with a 32P-labeled SR oligoprobe chosen in exon 4 of sCT I mRNA, which crosshybridized withexon 4 of sCT IV. Two hybridization signals of "600 bp were observedfor lanes 7 and 8. Hybridization was controlled by depositing, alongwith the different samples, a DNA fragment obtained from amplifi-cation ofUB cDNA with a set of primers specific for the sCT I mRNA(lane 1). (A is enlarged compared with B.)

The second round of PCR, using the same 5' primer and a3' primer specific for sCT IV, provided DNA fragments of theexpected size, 200 bp, which also hybridized to the aforemen-tioned SR oligo probe (Fig. 3 A and B). These results highlysuggested presence of RNA coding for sCT IV in salmon UBand brain. Identity of the amplified DNA fragment with thesCT IV genomic sequence was confirmed by sequencing. In theterminal peptide, situated after the cleavage amidation site, adifference with the genomic sequence of 1 nt (A instead of G)was noted in the first amino acid, inducing change of Gly to Ser.

Similarly, amplification of UB and brain cDNA, primarilywith a 5' primer specific for the sCGRP IV genomic sequenceand a 3' primer corresponding to the oligo d(T)17 box primer,and subsequently with an internal 3' primer, oligo CG4,resulted in a DNA fragment of "150 bp for each tissue (Fig.4). This fragment corresponded to the anticipated size and sowas in favor of the presence ofRNA coding for the sCGRP IVin the tested tissues. As PCR-amplified product was detectedon agarose gel only after the second round of PCR, sequencewas established on that DNA fragment. These results estab-lished that mRNAs coding for sCT IV and sCGRP IV werepresent in the salmon tissues, UB, and brain.

1 2 3 4 5 6 7 8

A B

FIG. 3. (A) Southern analysis of DNA fragments generated by anew round of amplification of PCR products of Fig. 2, using as 3'primer an internal oligo specific for sCT IV (oligo S4; Table 2), andthe same 5' primer as for the first round. Lanes 2-5, PCR controls; lane6, size marker; lanes 7 and 8, PCR products obtained with UB cDNAsand brain cDNAs, respectively; and lane 1, hybridization controlidentical to that in Fig. 2. A major band of the expected size ("200 bp)was detected in the two amplified tissues. (B) Autoradiography of theSouthern blot hybridized with the 32P-labeled SR oligo probe de-scribed in Fig. 2. Two signals were observed (lanes 7 and 8) corre-sponding to the agarose bands.

12346 Cell Biology: Jansz et al.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 14

, 202

2

Proc. Natl. Acad. Sci. USA 93 (1996) 12347

1 2 3 4

FIG. 4. Southern blot analysis of DNA fragments obtained afteramplification ofUB and brain cDNAs with specific primers for sCGRPIV. In a first round of PCR, a 5' primer specific of CGRP IV (oligoCG'4; Table 2) and a 3' primer corresponding to the end of thepoly(dT) box primer (oligo 41; Table 2) were used. No DNA fragmentwas UV detectable in these conditions (data not shown). A secondround of amplification was performed from these PCR products usinga 3' primer, an internal oligo specific for CGRP IV (oligo CG4; Table2), and the same 5' primer as that in the first round. In that case, twoDNA fragments of the expected size ("150 bp) were observed (lanes2 and 3), which were subsequently submitted to sequencing. Lane 1,PCR control; and lane 4, size marker.

Alignment and Comparison ofsCT IVAmino Acid Sequencewith the Established CT Amino Acid Sequences. Comparisonof the sCT IV amino acid sequence with those of the othersalmon isohormones indicated a difference of 5 aa with CT I,3 aa with CT II, and 4 aa with the CT III. Extension of thecomparison with amino acid sequences of the other nonmam-malian vertebrates showed that a greater homology exists withthe eel CT sequence than with the other sCT and, to a lesserextent, with the chicken CT, as they differed by 2 and 4 aa,respectively (Fig. 5). We must note that the 2-aa differencebetween the sCT IV and eel CT corresponded to amino acidsin position 15 and 22. These two, modified in sCT I, are alsofound in sCT II and sCT III.Comparison of sCGRP IV Amino Acid Sequence with the

Other Known CGRP Amino Acid Sequences. Alignment of thedifferent CGRP amino acid sequences indicated that sCGRPIV differed by 1 aa with the sCGRP I and by 4 aa with thechicken CGRP. Comparison of sCGRP IV with the humanmolecules showed a greater homology between the sCGRP IVmolecule and the human CGRP II molecule than between thesCGRP I molecule and the human CGRP II molecule (Fig. 6).

Phylogenetic Analysis. The phylogenetic tree confirmedevolutionary proximity of the sCT IV molecule with the eel CTmolecule and, to a lesser extent, with the chicken molecule(Fig. 7).

DISCUSSIONWe have identified in salmon genomic DNA of 0. keta a newCT gene encoding both for CT and CGRP molecules and

sCT I C S N L S T C V L G K L S Q E L H K L Q T Y P R T N T G S G T P

sCT I I - - - - - - - - - - - - - - D - - - - - - F - - - - - - A - V -

sCT III - - - - - - - M - - - - - - D - - - - - - F - - - - - - A -V -

sCT IV - - - - - - - - - - - - - - D - - - - - - F - - - DV - A - - -

eelCT - - - - - - - - - - - - - - - - - - - - - - - - - DV - A - - -

goldCT - - S - - - - - - - - - - - - - - - - - - - - - - - V - A - - -

sardCT - - - - - - - A - - - - - - - - - - - - S- - - - - V - A - - -

chicCT -A S - - - - - - - - - - - - - - - - - - - - - - D V - A - - -

rayCT - T S - - - - - V - - - - - Q - - - - - N I Q - - D - - A A - -

*CGRP I A C N T A T C V T H R L A D F L N R S G G N G N S N F V P T N V G A K A F

SCGRPIV - S-

chicCGRtP - - - - - - - - - - - - - - - - S - - - - V -K N - - - - - - - - S - - -

ranaCGRP - - - - - - - - - - - - - - - - S - - - - -A- N - - - - - - - - S - - -

hCGRP I - - D - - - - - - - - - - G LG - S- - - V V K N - - - - - - - - S ---

hCGRP II - - - - - - - - - - - - - G L -S - - - - - V - - - - - - - - - S -

rCGRP I S - - - - - - - - - - - - G L S - - - - V V K D - - - - - - - - S E -

rCGRPII S - - - - - - - - - - - - G L -S - - - - V V K D - - - - - - - - S

FIG. 6. Alignment of CGRP amino acid sequences. sCGRP I (5)and IV, chicken CGRP (chicCGRP; ref. 4), Rana CGRP (ranaCGRP;ref. 21), human CGRP I and II (hCGRP; ref. 22), and rat CGRP I andII (rCGRP; ref. 23).

demonstrated that this gene is transcribed in salmon tissues,UB, and brain of the pink salmon 0. gorbuscha. So, it appearsthat in salmon, the two transcription units further identifiedboth encoded CT and CGRP molecules, contrary to thosefound in human species, where only the gene I can generate CTtranscripts, the gene II being a pseudogene for human CT. Itremains to be established whether the transcription unitscoding for the sCT II and sCT III also coded for two CGRPmolecules.Up until now, three isohormones of CT have been reported

for the different salmon species studied: sCT I, sCT II, and sCTIII. Pink salmon, like sockeye and chum salmon, do possesssCT I and sCT II, while coho has sCT II and sCT III (8). Ourresults established that the pink salmon, 0. gorbuscha, couldexpress, in addition to sCT I and sCT II, another isohormone,the sCT IV. Comparison of the predicted sCT IV amino acidsequence with those of the other bony fish CT, confirmed thehigh susceptibility to modification of the amino acids situatedin positions 15 and 22 and also in positions 26, 27, and 29. Amajority of these modifications are maintained with evolution,as they are also found in another nonmammalian species, the

FIG. 5. Alignment of fish and chicken CT amino acid sequences.sCT I, II, III, and IV (8), eel CT (eelCT; ref. 16), goldfish CT (goldCT;ref. 17), sardine CT (sardCT; ref. 18), ray CT (rayCT; ref. 19), andchicken CT (chicCT; ref. 20).

FIG. 7. Phylogenetic tree of CT amino acid sequences of all knownspecies. SCT IV clustered with eel CT, while sCT I was on a separatebranch, and sCT II clustered with sCT III on an other bifurcation.Numbers on branches refer to bootstrap probabilities.

Cell Biology: Jansz et al.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 14

, 202

2

Proc. Natl. Acad. Sci. USA 93 (1996)

chicken. Comparison with a cartilaginous fish, the ray species,shows a low degree of homology between the CT amino acidsequences. This divergence seems to appear concomitantlywith the emergence of a calcified skeleton.To date, in human species, only one CT molecule has been

shown to be expressed, and a lower vertebrate CT-like mole-cule has been suggested (24). In salmon fish, the situation isvery different, as three molecules of CT have been observed(8), and we have established the possible expression of a fourthone. Existence of multiple isoforms of CT in salmonids may beexplained by the genomic tetraploidization event they under-went, leading to the presence of extensive gene duplication,and also by the persistence during evolution of a productive CTsequence. In addition to the presence of such multiple isoformsin salmon, we recently reported that different mRNA pathwaysmay be used to produce the sCT I molecule (25). Thesedifferences between human and salmon species are in favor ofan important physiological role for the peptide in salmonspecies, a role that would have disappeared during evolution.The higher biological activity of the salmon molecule com-pared with that of mammals (9) is consistent with this hypoth-esis. We and others previously reported important variationsof CT during the salmon migration and suggested a role for thepeptide during reproduction (14, 26). A CT action on plasmacalcium concentration of fishes has been highly suggested, butto date, no consensus has arisen about CT effect on calcemia(27-29). We recently established that, in addition to its endo-crine action, CT can exert a paracrine or autocrine role in thegill of these salmon species (30), and so, may locally performan important action in calcium regulation, as suggested by earlyin vitro experiments (31).Amino acid sequence analysis of the sCGRP IV indicates

that the molecule has a greater homology with the chickenCGRP sequence than has the sCGRP I molecule. Comparisonwith CGRP amino acid sequences of higher vertebratesshowed a somewhat greater homology with the human CGRPII. The small variations observed in the sequences of theCGRP molecules between salmon fishes and mammalianspecies imply important conservation of the molecule duringevolution and are consistent with the physiological importanceof the peptide.We previously demonstrated existence of sCGRP mRNA in

salmon tissues (32). By PCR, we established that a mRNApopulation coding for the sCGRP IV molecule is expressed inbrain tissue, as is the case for the CGRP I molecule (data notshown). Identification of CGRP-like molecules in fish brainhas been previously reported (32, 33). Such a localization ofthe peptide suggests that, as in mammals, CGRP can exert aneuropeptide role in fish.

Moreover, our study indicates that the sCT IV peptide canpotentially be expressed in UB and brain. Though we don'tquantify expression of the two peptides in tissues, efficiency ofamplification of the sCT IV in UB and brain is in favor of atissue-specific expression of the peptides, as it occurs inmammals (1).The similarity of this new sCT with the highly biologically

active eel CT and sCT II (9) allows to anticipate for animportant activity of the sCT IV and thus prompts us to furtherinvestigate the activity of the peptide. This will represent thenext step of this work, and it will involve biological studies andalso characterization of the sCT IV receptors. Similarly, itwould be important to compare activity of sCGRP IV to thatof human CGRP for clinical applications.

In conclusion, the identification of a new sCT gene is highlyimportant, as it may yield information on the evolution of the CTgene family, and also because of potential therapeutic interest inthe new CT' and CGRP molecules encoded by this gene.

We have appreciated the generous encouragement of the late

Robert 0. Peterson. The field work was supported in part by the R. P.Foundation. We thank Dr. P. Holthuizen for synthesizing the oligosused for PCR amplification of genomic DNA.

1. Amara, S. G., Jonas, V., Rosenfeld, M. G., Ong, E. S. & Evans,R. M. (1982) Nature (London) 298, 240-244.

2. Rosenfeld, M. G., Mermod, J. J., Amara, S. G., Swanson, L. W.,Sawchenko, P. E., Rivier, J.' Vale, W. W. & Evans, R. M. (1983)Nature (London) 304, 129-135.

3. Steenbergh, P. H., Hoppener, J. W. M., Zandberg, J., Van deVen, W. J. M., Jansz,. H. S. & Lips, C. J. M. (1984) J. Clin.Endocrinol. Metab. 59, 358-360.

4. Minvielle, S., Cressent, M., Lasmoles, F., Julienne, A., Milhaud,G. & Moukhtar, M. S. (1987) FEBS Lett. 223, 63-68.

5. Jansz, H. S. & Zandberg, J. (1992) Ann. N.Y Acad. Sci. 657,63-69.

6. Steenberg, H., Hoppener, J. W. M., Zandberg, J., Lips, C. J. M.& Jansz, H. S. (1985) FEBS Lett. 183, 403-407.

7. Alevizaki, M., Shiraishi, A., Rasool, F. V., Ferrier, G. J. M.,Maclntyre, I. & Legon, S. (1986) FEBS Lett. 206, 47-52.

8. Niall, H. D., Keutmann, H. T., Copp, D. H. & Potts, J. T., Jr.(1969) Proc. Natl. Acad. Sci. USA 64, 771-778.

9. Guttmann, S. (1980) Excerpta Med. Int. Congr. Ser. 540, 11-24.10. Poschl, E., Lindley, I., Hofer, E., Seifert, J. M., Brunowski, W. &

Besemer, J. (1987) FEBS Lett. 226, 96-100.11. Chomcsynsky, P. & Sacchi, N. (1987) Anal. Biochem. 162, 156-

159.12. Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, R.,

Horn, G. T., Mullis, K. B. & Erlich, H. A. (1988) Science 239,487-491.

13. Sanger, F., Nicklens, S. & Coulson, A. R. (1977) Proc. Natl. Acad.Sci. USA 74, 5463-5467.

14. Maubras, L., Cressent, M., Minvielle, S., Taboulet, J., Julienne,A., Milhaud, G. & Moukhtar, M. S. (1990) Biochem. Biophys. Res.Commun. 173, 788-794.

15. Kanehisa, M. I. (1984) Nucleic Acids Res. 12, 203-213.16. Otani, M., Yamauchi, H., Meguro, T., Kitazawa, S., Watanabe,

T. X. & Orimo, H. (1976) J. Biochem. (Tokyo) 79, 345-352.17. Sasayama, Y., Ukawa, K., Kai-Ya, H., Oguro, C., Takei, Y.,

Watanabe, T. X., Nakajima, K. & Sakakibara, S. (1993) Gen.Comp. Endocrinol. 89, 189-194.

18. Suzuki, N., Nose, Y., Kase, Y., Sasayama, Y., Takei, Y., Naga-sawa, H., Watanabe, T. X., Nakajima, K. & Sakakibara, S. (1994)Zool. Sci. 11, 713-716.

19. Takei, Y., Takahashi, A., Watanabe, T. X., Nakajima, K., Saka-kibara, S., Sasayama, Y., Suzuki, N. & Oguro, C. (1991) Biol. Bull.180, 485-488.

20. Lasmoles, F., Julienne, A., Desplan, C., Milhaud, G. &Moukhtar, M. S. (1985) FEBS Lett. 180, 113-116.

21. Conlon, J. M., Tonon, M. C. & Vaudry, H. (1993) Peptides 14,581-586.

22. Steenbergh, P. H., Hoppener, J. W. M., Zandberg, J., Visser, A.,Lips, C. J. M. & Jansz, H. S. (1986) FEBS Lett. 209, 97-103.

23. Amara, S. G., Arriza, J. L., Leff, S. E. M., Swanson, L. W., Evans,R. M. & Rosenfeld, M. G., (1985) Science 229, 1094-1097.

24. Fischer, J. A., Tobler, P. H., Henke, H. & Tschopp, F. A. (1983)J. Clin. Endocrinol. Metab. 57, 1314-1316.

25. Martial, K., Maubras, L., Taboulet, J., Julienne, A., Milhaud, G.,Moukhtar, M. S. & Cressent, M. (1994) Gene 149, 277-281.

26. Watts, E. G., Copp, D. H. & Deftos, L. J. (1975) Endocrinology96, 214-218.

27. Chan, D. K. O., Jones, I. C. & Smith, R. N. (1968) Gen. Comp.Endocrinol. 11, 243-245.

28. Pang, P. K. T. (1971) J. Exp. Zool. 178, 89-100.29. Yamauchi, H., Matsuo, M., Yoshido, A. & Orimo, H. (1978) Gen.

Comp. Endbcrinol. 34, 343-346.30. Martial, K., Maubras, L., Taboulet, J., Julienne, A., Berry, M.,

Milhaud, G., Benson, A. A., Moukhtar, M. S. & Cressent, M.(1994) Proc. Natl. Acad. Sci. USA 91, 4912-4914.

31. Milhaud, G., Bolis, B. & Benson, A. A. (1980) Proc. Natl. Acad.Sci. USA 77, 6935-6936.

32. Maubras, L., Taboulet, J., Pidoux, E., Lasmoles, F., Julienne, A.,Milhaud, G., Benson, A. A., Moukhtar, M. S. & Cressent, M.(1993) Peptides 14, 977-981.

33. Fouchereau-Peron, M., Arlot-Bonnemains, Y., Taboulet, J., Mil-haud, G. & Moukhtar, M. S. (1990) Regul. Pept. 27, 171-179.

12348 Cell Biology: Jansz et al.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 14

, 202

2