24 S. WATABE ET AL.

© 2000 WILEY-LISS, INC.

Complete Amino Acid Sequence of Mytilus AnteriorByssus Retractor Paramyosin and Its PutativePhosphorylation Site

SHUGO WATABE,* KEIKO IWASAKI, DAISUKE FUNABARA,YASUSHI HIRAYAMA, MISAKO NAKAYA, AND KIYOSHI KIKUCHILaboratory of Aquatic Molecular Biology and Biotechnology, GraduateSchool of Agricultural and Life Sciences, University of Tokyo,Tokyo 113-8657, Japan

ABSTRACT A cDNA encoding the full-length paramyosin molecule was cloned from the musselMytilus galloprovincialis, a species closely related to Mytilus edulis. It contained 3,497 nucleotides(nt), with 79 and 826 nt for the 5′ and 3′ non-coding regions, respectively. The coding region wascomposed of 2,592 nt for 864 amino acid residues, a size typical of paramyosin. While genomic DNAdigests with either HindIII or PstI exhibited a single band when hybridized with a SacI fragment ofparamyosin cDNA, the digests with either EcoRV or EcoRI showed two bands, suggesting that themussel has at least two genes encoding paramyosin. The mRNAs encoding paramyosin were mostabundant in muscle tissues from byssus retractor and adductor muscles. Only traces of paramyosintranscripts were found in the tissue of foot, gill, inner mantle, and outer mantle. The samephosphorylatable peptide previously reported for paramyosin from the bivalve Mercenaria mercenaria,Ser-Arg-Ser-Met-Ser(P)-Val-Ser-Arg (Watabe et al. 1989. Comp Biochem Physiol 94B:813–821) wasfound in the C-terminal non-helical part of this Mytilus paramyosin. We predict that this particularparamyosin has a coiled-coil structure composed of two α-helices that show the heptad repeats (a-b-c-d-e-f-g) with further 28-amino acid repeat zones, where a and d tend to be occupied by nonpolarresidues. J. Exp. Zool. 286:24–35, 2000. © 2000 Wiley-Liss, Inc.

Paramyosin from the molluscan adductor mus-cle was first reported as tropomyosin A and laterreceived the now prevalent name paramyosin(Bailey, ’56). It has a fibrous structure as commontropomyosin. However, paramyosin is not solublein a low-ionic-strength buffer at neutral pH as myo-sin and its C-terminal proteolytic products, rod, andlight meromyosin, whereas tropomyosin is solubleunder corresponding conditions (Kominz et al., ’58;Johnson et al., ’59; Woods and Pont, ’71). A widedistribution of paramyosin now is recognized in themuscles of invertebrates as diverse as nematodes,arthropods, and echinoderms (Elfvin et al., ’76;Winkelman, ’76; SriKantha et al., ’90); but is notfound in vertebrates.

Paramyosin is a coiled-coil protein of two α-heli-ces (Cohen et al., ’87) that form the core of thickfilaments beneath myosin surface (Epstein et al.,’76; Deitiker and Epstein, ’93). The invertebratesmooth muscles such as adductor muscles ofbivalves, and especially the anterior byssus retrac-tor muscle (ABRM) of the mussel, that sustainstrong tension with little energy expenditure are

called catch (Baguet, ’68; Baguet and Gillis, ’68).Such catch muscles have extraordinary large thickfilaments consisting of myosin and paramyosin>50 nm in diameter that enable such energy-sav-ing muscle contraction (Lowy et al., ’64; Cohen etal., ’71; Szent-Gyorgyi et al., ’71; Nonomura et al.,’74; Levine et al., ’76; Hauck and Achazi, ’87;Matsuno et al., ’96). This tension is developed atdecreased concentrations of Ca2+ following activecontraction triggered by Ca2+ (Baguet and Gillis,’68; Marchand-Dumont and Baguet, ’75). Para-myosin-containing filaments were claimed asfused to each other in catch state, thus producinghigh tension with little energy expenditure (John-son, ’62).

Grant sponsor: Ministry of Education, Science, Sports, and Cul-ture of Japan.

*Correspondence to: Shugo Watabe, Laboratory of Aquatic Molecu-lar Biology and Biotechnology, Graduate School of Agricultural andLife Sciences, University of Tokyo, Tokyo 113-8657, Japan. E-mail:awatabe@mail.ecc.u-tokyo.ac.jp

Received 14 July 1998; Accepted 9 April 1999

JOURNAL OF EXPERIMENTAL ZOOLOGY 286:24–35 (2000)

MYTILUS PARAMYOSIN 25

Paramyosin isolated from the bivalve Mer-cenaria mercenaria has a molecular weight of220,000, is shaped 135 nm long and 2 nm in di-ameter, and consists of two intertwined α-helicalpolypeptide chains (Cowgill, ’72, ’74, ’75a,b). It issusceptible to proteolysis, easily losing a C-ter-minal region during preparation (Stafford et al.,’72). Intact paramyosin, called α-type, is appar-ently 5,000 times greater in subunit molecularweight than that extracted by a conventional pro-cedure, β-paramyosin, without reducing agents orprotease inhibitors. β-Paramyosin is soluble overthe range of ionic strength, in which the solubil-ity of α-paramyosin rapidly fluctuates (Yeung andCowgill, ’76; Edwards et al., ’77; Merrick andJohnson, ’77).

While serotonin mediates release of catch in theABRM of Mytilus edulis, accompanying accumu-lation of cAMP (Achazi et al., ’74; Twarog et al.,’77), Achazi (’79) found that paramyosin from theABRM could be phosphorylated by endogenouskinase(s) and by exogenous cAMP-dependent pro-tein kinase. There is a strong evidence for a roleof cAMP-dependent protein kinase in relaxationfrom catch. Pfitzer and Ruegg (’82) found thatcatch was released in skinned ABRM fibers ei-ther by cAMP or by the catalytic subunit of cAMP-dependent protein kinase. However, paramyosinin the ABRM activated with acetylcholine andthus in a phasic contraction, or catch state, wasmore phosphorylated than that in the fiber re-leased from catch (Achazi, ’79). These results sug-gest that cAMP accumulation and its dependentprotein kinase does not necessarily act on para-myosin when catch is released, but rather thatparamyosin in intact muscle is dephosphorylatedin this relaxing state. Regardless of such a con-fused interpretation on paramyosin phosphoryla-tion and dephosphorylation transition in catchmuscles, Cohen (’82) proposed that the interac-tion of paramyosin with myosin subfragment 2 isenhanced on paramyosin phosphorylation, possi-bly correlating with slow cross-bridge cycling tofacilitate catch.

Among the Mytilus species investigated previ-ously, Cooley et al. (’79) found phosphorylation ofparamyosin from Mytilus edulis ABRM. Subse-quently, Radlick and Johnson (’82) claimed in abrief note that only α-paramyosin could be phos-phorylated by endogenous kinase(s) and, there-fore, phosphorylation site(s) would be located nearthe C-terminus. We also have observed thatparamyosin in adductor myofibrils of M. mer-cenaria was phosphorylated with endogenous

kinase(s) and that isolated paramyosin could bephosphorylated by cAMP-dependent protein ki-nase (Watabe et al., ’89; Watabe and Hartshorne,’90; Watabe et al., ’90). On the other hand,Castellani and Cohen (’87a,b) found that the myo-sin heavy chain could be phosphorylated in ABRMmyofibrils by endogenous protein kinase(s). In ad-dition, it has been suggested that phosphoryla-tion of a light chain in scallop smooth musclemyosin is important in the catch contraction(Sohma et al., ’85), and a light chain kinase, notdependent on calmodulin, was isolated (Sohmaand Morita, ’86). Very recently, Siegman et al. (’97,’98) discovered that relaxation from catch accom-panies phosphorylation of a twitchin-related pro-tein at serotonergic nerve stimulation in intactMytilus ABRM and application of cAMP in per-meabilized muscles, but nothing significant withparamyosin, myosin heavy chain, and myosin lightchains. They have also shown that phosphoryla-tion of twitchin-related protein controls force pro-duction by modulating calcium sensitivity underconditions not historically considered to be asso-ciated with catch (Siegman et al., ’98).

It is evident that whether paramyosin and itsphosphorylation also regulate catch contractionand relaxation cycle remains an interesting ques-tion. To address such a question, it seems prima-rily important to clarify more details on thestructural properties of paramyosin. Thus, the ob-jective of the present study was to clone a cDNAencoding paramyosin from Mytilus ABRM in or-der to reveal its primary structure, a putativephosphorylation site, and tissue distribution.

MATERIALS AND METHODSMaterials

Live specimens of the mussel Mytilus gallo-provincialis (average body weight including theshell, 24 g) cultured at the coastal areas of MiyagiPrefecture in the Pacific Ocean, the northeasternregion of the Honshu Island of Japan (Inoue et al.,’95), were obtained at the Tokyo Central WholesaleMarket. The ABRM pieces were carefully dissectedfrom the mussel while avoiding contamination withvisceral parts, immediately frozen with liquid ni-trogen, and stored at –80°C until use. Several tis-sues other than ABRM were also isolated and usedfor Southern and Northern blot analyses.

Isolation of paramyosinα-Paramyosin was isolated using 19 g ABRM,

according to the procedure of Merrick and Johnson

26 S. WATABE ET AL.

(’77), except EDTA was omitted (Watabe et al.,’89). Isolated paramyosin was further purified byhigh-speed gel filtration using a TSK G3000SWGcolumn (0.75 × 60 cm) (Tosoh) equilibrated with0.1 M potassium phosphate (pH 6.8) containing0.5 M KCl and 0.1 mM DTT.

ElectrophoresisSDS-polyacrylamide gel electrophoresis (SDS-

PAGE) was performed using gradient (7.5–20%)slab gels with the discontinuous buffer systemsof Laemmli (’70). After electrophoresis, proteinswere stained with 0.1% Coomassie Brilliant BlueR-250 and destained with a solution containing25% methanol and 7% acetic acid. Paramyosincontent in myofibrils was determined by densito-metric analysis. Protein molecular weight mark-ers were purchased from Sigma: rabbit musclemyosin heavy chain (205 kDa), β-galactosidase(116 kDa), rabbit muscle phosphorylase b (97.4kDa), bovine serum albumin (66 kDa), ovalbumin(45 kDa), and carbonic anhydrase (29 kDa).

N-terminal amino acid sequencingLimited proteolysis of paramyosin dissolved in

0.1 M sodium phosphate (pH 7.0) was performedat 30°C using lysyl endopeptidase (Achromobacterprotease I, Wako) with an enzyme-to-paramyosinweight ratio of 1:500.

The N-terminal amino acid sequences of para-myosin peptides were determined by the methodof Matsudaira (’87) as follows. Proteins separatedon SDS-PAGE were electrically transferred ontoImmobilon polyvinylidene difluoride membranes(Millipore) and stained with Coomassie BrilliantBlue R-250. The parts of the membrane carryingthe blotted proteins were cut out with a clean ra-zor. Several membranes were placed together ina blot cartridge block of a Perkin Elmer/AppliedBiosystems model 476A protein sequencer with anon-line data processor system, model 610A. Thesequence homology search was performed usingthe Swiss-Prot database coordinated with the In-herit program (Perkin Elmer/Applied Systems).

Construction of cDNA LibraryTotal RNA was prepared from 7 g of the ABRM

by the method of Chomczynski and Sacchi (’87).Poly (A)+ mRNA was isolated from total RNA us-ing an oligo(dT)-cellulose column (Pharmacia)(Sambrook et al., ’89). Five micrograms of mRNAwere used for double-stranded cDNA synthesiscarried out using Pharmacia cDNA synthesis kitswith an oligo dT primer. cDNA library was con-

structed using synthesized cDNAs in a phage vec-tor λZAP II, according to supplier’s instructions(Stratagene), after ligating cDNA to a NotI/EcoRIlinker.

cDNA cloningThe cDNA fragment that encodes a part of the

ABRM paramyosin molecule was obtained by PCR(Saiki et al., ’88). DNA primers for PCR were de-signed referring to N-terminal amino acid se-quences of ABRM, together with data reported forparamyosins from other species. The PCR ampli-fications were performed for 1 min at 94°C fordenaturation, 1.5 min at 50°C for annealing, and2.0 min at 72°C for extension using a DNA ther-mal cycler model 9600 and GeneAmp kits (PerkinElmer Cetus Instruments). This procedure wascarried out for 30 cycles and the last extensionstep performed at 72°C for 5 min. The 100 µl re-action mixture contained 200 nM dNTPs, 50 pmolof forward and reverse primers, 2 units of TaqDNA polymerase, and 50 ng of first-strand cDNA.

An approximately 500 bp DNA fragment ob-tained by PCR was randomly labeled with a DIGDNA labeling and detection kit (BoehringerMannheim), according to the manufacturer’s in-struction. This fragment was used to screenABRM cDNA library. Hybridization was carriedout at 68°C in 5× SSC (saline sodium citrate, 1×concentration containing 0.15 M NaCl and 15 mMsodium citrate, pH 7.0) containing 1.0% blockingreagent, 0.1% N-lauroylsarcosine, and 0.02% SDS.

DNA sequence analysiscDNA restriction fragments of paramyosin were

subcloned into plasmid pBluescript SK– using Es-cherichia coli strain MV1190 as a host bacterium.The plasmid DNAs were purified by an alkalinelysis method (Sambrook et al., ’89) and used forDNA sequence analysis. Sequencing was per-formed for both 5′ and 3′ strands of subclones la-beled with Dye Deoxy terminator cycle sequencekits using a DNA sequencer model 373A (PerkinElmer/Applied Biosystems).

Southern blot analysisHomogenizing the hepatopancreas and subse-

quently treating with proteinase K (Sambrook etal., ’89) isolated genomic DNA. For Southern blotanalysis, 10 µg of genomic DNAs were digestedwith a treatment series with restriction endonu-cleases and electrophoresed in 0.7% agarose gels.The gels were processed with slight modifications,based on Sambrook et al. (’89), denatured with

MYTILUS PARAMYOSIN 27

0.5 M NaOH containing 1.5 M NaCl, and trans-ferred to HybondN+ nylon membranes (Amer-sham), omitting renaturing steps, and weresubsequently baked at 80°C for 15 min. The mem-branes were hybridized at 65°C in 0.5 M sodiumphosphate (pH 7.2) containing 1 mM EDTA and7% SDS with a 32P-labeled DNA fragment ob-tained by SacI digestion of paramyosin cDNA andsubsequently randomly primed in the presence of[α-32P]dCTP (Church and Gilbert, ’84). Membranefilters were washed at 65°C with several bufferchanges of decreasing SSC concentrations from 5×to 0.1× and autoradiographed on X-ray films withintensifying screens at –80°C.

Northern blot analysisThe following procedures were performed essen-

tially according to Sambrook et al. (’89). Total RNAwas isolated from tissues using an Isogen solu-tion (Nippon Gene). The concentration of RNA wasdetermined spectrophotometrically by reading A260for each sample before gel electrophoresis. Tenmicrograms of total RNAs isolated from varioustissues of the mussel were denatured at 65°C for15 min in 50% formamide, subjected to electro-phoresis on a 0.9% agarose gel in 0.04 M 3-(N-morpholino)propanesulfonic acid containing 0.44M formamide, 0.01 M sodium acetate, and 1 mMEDTA. Gels were stained with ethidium bromideand photographed before transfer to HybondN+ ny-lon membranes to ascertain that an equal amountof intact RNA blots were hybridized. The mem-branes were air-dried and baked at 80°C for 15min prior to hybridization. The subsequent pro-cedures for hybridization and autoradiographywere carried out as those described for Southernblot analysis.

RESULTSIsolation of paramyosin and N-terminal

amino acid sequencingParamyosin was prepared by a conventional

method to be pure enough for subsequent proteolyticexperiments. When several bands produced bylimited proteolysis with lysyl endopeptidase weresubjected to N-terminal amino acid sequencinganalysis, one fragment gave the sequence of 30 resi-dues (Fig. 1). When this sequence was comparedwith the primary structure of tapeworm Echinococ-cus granulosus paramyosin (Muhlschlegel et al., ’93),which provided the highest score in homologysearch, the sequence obtained was best alignedwith that of tapeworm paramyosin from the 155th

to 184th residue from the N-terminus. The sequenceidentity between the Mytilus ABRM fragment andthe corresponding region of tapeworm paramyosinwas 53%, suggesting that the paramyosin of ourinterest has a primary structure considerably dif-ferent from those of other paramyosins.

cDNA cloning of ABRM paramyosinThe first attempt to use the DNA nucleotide se-

quences available from other paramyosins asprimers was unsuccessful. From the above N-ter-minal amino acid sequence, we newly designedseveral mix primers capable of amplifying DNAfragments encoding our paramyosin. One exampleof primer sets used was 5′-dGCGGATCCATGTC-(A/G/T/C)GC(A/G/T/C)GA(T/C)TC(A/G)AA(A/G)AT-(A/T/C)GA-3′ of the 5′-mixed primer designed fromthe determined sequence, MSADSKI (Fig. 1), and5′-dATGGATCC(T/C)TT(A/G/T/C)AG(T/C)TC(T/C)TC(A/G/T/C)AC(T/C)TC(T/C) TC-3′ of the 3′-mixed primer degenerated from the conserved se-quence for paramyosins reported so far that

Fig. 1. Lysyl endopeptidase digestion of Mytilus anteriorbyssus retractor muscle paramyosin for N-terminal amino acidsequence determination. SDS-PAGE (A) was carried out forparamyosin isolated from ABRM (lane 1) and subjected tolysyl endopeptidase digestion for 30 min (lane 2). A fragmentshown in lane 2 with arrowhead was successfully determinedfor its N-terminal amino acid sequence (B). PM, paramyosin;aPM, aggregated paramyosin.

28 S. WATABE ET AL.

corresponds to EEFEELK of tapeworm para-myosin (Muhlschlegel et al., ’93) (Fig. 2). Bothprimers contained BamHI recognition sequence tofacilitate subsequent subcloning. When the de-duced amino acid sequence from an amplifiedDNA fragment was compared with that of para-myosin from the tapeworm E. granulosus (Muhl-schlegel et al., ’93), the two sequences showed 49%identity (Fig. 2).

The above PCR product was used for screeningcDNA clones that encode paramyosin in the library

constructed from the ABRM, giving one clone withabout 3.5 kbp. When primers of Bluescript wereused to determine DNA nucleotide sequences in the5′ and 3′ sites after subcloning, the initiation codon,ATG, and poly A tail were found in respective de-terminations. Therefore, it was considered that thisclone encoded the full length of the paramyosin mol-ecule. The clone was found to contain four SacI andthree HincII sites, which facilitated the determina-tion of the whole DNA nucleotide sequence by us-ing such sites for subcloning (Fig. 3). Although other

Fig. 2. Comparison of the deduced amino acid sequenceof a part of the paramyosin molecule of Mytilus anterior bys-sus retractor muscle with that of Echinococcus paramyosin.An underline and a double underline indicate the deducedamino acid sequences from the DNA fragment of MytilusABRM amplified by PCR that were matched with an N-ter-

minal amino acid sequence for the Mytilus ABRM paramyosinfragment shown in Fig. 1 and that used for regeneration ofthe 3′-primer from data previously reported for paramyosinsfrom other species, respectively. Identical amino acids areshown by periods. The sequence of paramyosin for Echino-coccus granulosus was cited from Muehlschlegel et al. (’93).

Fig. 3. Restriction endonuclease maps of cDNA clones en-coding paramyosin of Mytilus anterior byssus retractormuscle. Subcloned were regions including 4 digests of SacIand 3 digests of HincII digests indicated by open bars. The

SacI-digested DNA fragment indicated as a solid bar (1,010–1,650 nt) was applied to Southern and Northern blot analy-ses (see Figs. 6 and 7).

MYTILUS PARAMYOSIN 29

efforts resulted in the isolation of several para-myosin clones, these were all shorter than the aboveclone, lacking the 5′ site (data not shown). As faras all these clones were analyzed, only the 3,358thbase in the 3′ non-coding region, thymine, was sub-stituted for adenine.

Complete primary structure ofABRM paramyosin

The complete DNA nucleotide sequence and de-duced amino acid sequence of ABRM paramyosinare shown in Fig. 4. The nucleotide sequences ap-pear in the DDBJ/EMBL/GenBank nucleotide se-quence databases with the accession numberAB016070. ABRM paramyosin was composed of864 amino acid residues, giving a molecularweight of 99,568. When this amino acid sequencewas compared with those of other paramyosinsreported so far, the highest identity was only 54%with the trematode Schistosoma mansoni (Lac-lette et al., ’91), whereas the lowest value of 39%was obtained in comparison with the fruitflyDrosophila melanogaster (Vinos et al., ’92) (Table1). Since paramyosin is markedly species-specificin terms of its primary structure, it seems rea-sonable that our first attempt to use previouslyreported DNA nucleotide sequences for PCR prim-ers was unsuccessful.

Figure 5 shows the ABRM paramyosin sequencearranged in 28-residue repeat zones with extraskip residues inserted at five places. N- and C-terminal sequences containing 28 and 19 aminoacid residues, respectively, were supposed to benon-helical. Interestingly, the sequence reportedto be phosphorylatable and located at the extremeC-terminal region for M. mercenaria paramyosin,Ser-Arg-Ser-Met-Ser(P)-Val-Ser-Arg (Watabe etal., ’89), was found in the ABRM C-terminal non-helical part. This identity strongly indicates thatphosphorylation of ABRM paramyosin occurs atthe same site.

Southern blot analysisGenomic DNA was obtained from the hepatopan-

creas of M. galloprovincialis and digested with vari-ous endonucleases that included EcoRI, EcoRV,HindIII, and PstI. Digests were size-fractionated byagarose-gel electrophoresis and transferred ontonylon membranes, followed by hybridization witha 32P-labeled DNA fragment obtained by SacI di-gestion of paramyosin cDNA (Fig. 3). This fragmentencoded the region from the 1,010 to 1,650 nucle-otide (nt) of the cDNA shown in Fig. 4. GenomicDNA digested with either HindIII or PstI contained

a single band, whereas EcoRI and EcoRV digestsexhibited two bands (Fig. 6). The occurrence ofparamyosin isoforms has been reported for D.melanogaster, which has a miniparamyosin isoformof 60-kDa subunit in addition to a common para-myosin of 100-kDa subunit (Maroto et al., ’95).Therefore, we amplified by PCR the genomic DNAfragment encoding the region, which correspondedto the cDNA probe used for Southern blot analysis.However, this fragment was not digested eitherwith EcoRI or EcoRV (data not shown). Since therewere no clones in the present study encodingABRM paramyosin, other than that shown in Fig.4 as stated before, it seems that Mytilus has mul-tiple copy genes that encode paramyosin.

Northern blot analysisTotal RNA was prepared from various tissues

of M. galloprovincialis, including ABRM, poste-rior byssus retractor muscle, adductor muscle, gill,foot, outer mantle, and inner mantle. Ten micro-grams each of RNA preparations were subjectedto agarose-gel electrophoresis under denaturingconditions and hybridized with the same DNAprobe as used for Southern blot analysis. The twolarger rRNAs (18S and 26S) (Collier, ’83), togetherwith commercially available 16S and 23S rRNAsfrom E. coli, were used as the molecular weightmarkers after staining with ethidium bromide.The mRNAs encoding ABRM paramyosin weremost abundant in muscle tissues from byssus re-tractor and adductor muscles (Fig. 7). Only traceamounts of paramyosin transcripts were found infoot, gill, outer mantle, and inner mantle, wherecontamination with muscle tissues was avoidedas much as possible.

DISCUSSIONCertain molluscan smooth muscles, including

the mussel ABRM, exhibit catch contraction wheretension is maintained with little energy expendi-ture. The catch state is resolved by stimulation ofrelaxing nerve fibers, when they release seroto-nin (Achazi et al., ’74; Twarog et al., ’77). Sincethe release of serotonin increases cAMP levels(Cole and Twarog, ’72; Achazi et al., ’74; Twarog,’76), this second messenger is implicated in theabolition of catch state.

Achazi (’79) reported that in Mytilus edulis thecontractile apparatus of catch muscles, ABRM, con-tains 3.5–10% of the proteins phosphorylatable withendogenous kinase(s). The two phosphorylated pro-teins had molecular masses of 295 and 106 kDa,the latter protein being identified as paramyosin.

30 S. WATABE ET AL.

Fig. 4. The complete nucleotide sequence of cDNA encod-ing paramyosin of Mytilus anterior byssus retractor muscle

and its deduced amino acid sequence. Polyadenylation sig-nals are underlined.

MYTILUS PARAMYOSIN 31

TABLE 1. Comparison of paramyosin amino acid sequence of Mytilus galloprovincialis paramyosinwith those from other paramyosins1

% IdentityM. galloprovincialis S. mansoni E. granulosus T. solium O. volvulus C. elegans

S. mansoni 54E. granulosus 53 75T. solium 53 72 96O. volvulus 41 34 33 33C. elegans 40 32 32 30 92D. melanogaster 39 35 35 30 49 471Paramyosins from other species were cited for Echinococcus granulosus (Muhlschlegel et al., ’93), Taenia solium (Landa et al., ’93), Schisto-soma mansoni (Laclette et al., ’91), Onchocerca volvulus (Dahmen et al., ’93), Caenorhabditis elegans (Kagawa et al., ’89), and Drosophilamelanogaster (Vinos et al., ’91).

Fig. 5. The amino acid sequence of paramyosin of Mytilusanterior byssus retractor muscle arranged in 28-residue re-peat zones. Extra skip residues are inserted in the patternat 113, 310, 507, 536, and 733. The positions d, e, f, g, a, b,and c of the seven-residue coiled-coil repeat are marked abovethe columns. Hydrophobic residues underlined usually occupy

a and d. N- and C-terminal regions in parentheses are notincluded in 28-residue repeat zones. The putative phos-phorylatable region that has the same sequence of phospho-rylated Mercenaria paramyosin (Watabe et al., ’89) isunderlined in the C-terminal non-helical region where aphosphorylatable amino acid, serine, is boxed.

32 S. WATABE ET AL.

These two proteins from the ABRM released fromcatch were more phosphorylated by exogenouslyadded cAMP-dependent protein kinase, suggestingthat the serotonin-mediated release of catch wasaccompanied by dephosphorylation of paramyosin.Such an inference is somewhat confusing as onecan deduce from the following interpretation. Accu-mulation of cAMP on release from catch impliesparticipation of cAMP-dependent protein kinasein phosphorylation of protein components in thecatch muscle, while paramyosin in this muscle isactually phosphorylated in vitro by cAMP-depen-dent protein kinase. Although myosin heavy (Cas-tellani and Cohen, ’87a,b) and light chains (Sohmaet al., ’85; Sohma and Morita, ’86) from molluscancatch muscles have been reported to be also phos-phorylated, paramyosin is still one of the candidateproteins that possibly regulate catch contraction andrelaxation by phosphorylation, in spite of such am-biguity. However, it should be noted that twitchin-related protein was only phosphorylated and noother proteins, including myosin heavy chains, myo-sin light chain, and paramyosin, were significantduring the release of catch at serotonergic nervestimulation in intact Mytilus ABRM and applica-tion of cAMP in permeabilized muscles (Siegmanet al., ’97, ’98). Therefore, functional significance of

Fig. 6. Southern blot analysis for genomic organizationof paramyosin of Mytilus anterior byssus retractor muscle.Genomic DNA was prepared from the hepatopancreas anddigested with various endonucleases that included EcoRI,EcoRV, HindIII, and PstI. A DNA fragment obtained by SacIdigestion of paramyosin cDNA (see Fig. 3) was 32P-labeledand used as a probe.

Fig. 7. Northern blot analysis for paramyosin from theanterior byssus retractor muscle of Mytilus galloprovincialisin various tissues. Total RNA was isolated from various tis-sues and subjected to electrophoresis by using denatured aga-rose gels (lower panel), capillary-transferred to HybondN+

nylon membranes and hybridized with 32P-labeled DNA probe(upper panel). The tissues examined were ABRM (lane 2),foot (lane 3), gill (lane 4), posterior byssus retractor muscle(lane 5), outer mantle (lane 6), inner mantle (lane 7), andadductor muscle (lane 8). rRNA markers from Escherichiacoli were contained in lane 1. Refer to the legend of Fig. 3 fora probe.

paramyosin phosphorylation in molluscan contrac-tion and relaxation cycle is still ambiguous, althoughits phosphorylation inhibited steady-state Mg2+-AT-Pase activity of M. mercenaria adductor myofibrils(Watabe et al., ’89).

The phosphorylation site has been determinedfor paramyosin from the molluscan catch muscleof the bivalve M. mercenaria (Watabe et al., ’89).Although its site has been assumed to be proxi-mal to the C-terminus of the molecule, the exactlocation has remained unidentified. In addition,the primary structure of molluscan paramyosinhas not yet been determined despite the availabil-ity of data on paramyosins for species belongingto other invertebrate groups such as nematodes andinsects (Kagawa et al., ’89; Limberger and Mc-Reynolds, ’90; Laclette et al., ’91; Vinos et al., ’91,’92; Dahmen et al., ’93; Landa et al., ’93; Muhl-schlegel et al., ’93; Nara et al., ’94). Therefore, wecarried out N-terminal amino acid sequence analy-sis for ABRM paramyosin after limited proteolysis.Using the determined sequence, together with thosefrom available data for other paramyosins, primersto amplify cDNA coding for ABRM paramyosin were

MYTILUS PARAMYOSIN 33

successfully designed that resulted in isolation of aclone encoding the full length of the paramyosinmolecule from Mytilus ABRM. Its deduced aminoacid sequence revealed that the ABRM C-terminalnon-helical part contained the peptide, Ser-Arg-Ser-Met-Ser(P)-Val-Ser-Arg, which has been re-ported to be phosphorylatable and located at theextreme C-terminal region for M. mercenariaparamyosin (Watabe et al., ’89). The SxSxA motifis found in the N-terminal non-helical region ofother paramyosin (Kagawa et al., ’89; Limbergerand McReynolds, ’90; Dahmen et al., ’93), whichis a possible target sequence for phosphorylationof Caenorhabditis elegans paramyosin (Schrieferand Waterson, ’89). However, the paramyosin ofMytilus galloprovincialis, reported in this study,contained no SxSxA motif, suggesting that thephosphorylation site is only located in the C-ter-minal non-helical region.

It is well known that the paramyosin moleculeis composed of two identical α-helices to give acoiled-coil structure. This structure has been mostextensively studied with myosin rod, which has acharacteristic regular pattern consisting of sevenamino acid residues, a, b, c, d, e, f, and g, wherehydrophobic amino acid residues are concentratedat alternate intervals of three and four residuesalong the length of the chain at positions a and d(Parry, ’81; McLachlan and Karn, ’82). The resi-dues at positions a and d form a closely packedhydrophobic interface between the two strands ofthe coiled-coil structure. The surface of the coiled-coil structure is highly charged, with acidic andbasic residues clustered mainly in the outer posi-tions b, c, and f. Myosin rod further forms a 28-residue unit with a repetitive sequence, when skipresidues are inserted at appropriate positions(McLachlan and Karn, ’82; Offer, ’90).

For most part of its length, paramyosin appearsto form an α-helical coiled-coil structure and showsthe expected heptad repeat of hydrophobic aminoacid residues and the 28-residue repeat of chargedamino acids (Cohen et al., ’87; Kagawa et al., ’89;Laclette et al., ’91; Vinos et al., ’92; Dahmen, ’93:Landa et al., ’93). The present paramyosin ofMytilus ABRM investigated in this study also hasthe same structure as those of other paramyosins,showing the expected heptad repeat of hydropho-bic amino acid residues and the 28-residue repeatof charged amino acids. A close resemblance be-tween the structures of paramyosin and the myo-sin rod is apparently essential for these proteinsto assemble together in muscle thick filaments(Cohen et al., ’87, ’98). It has been demonstrated

that a single amino acid substitution can alter theassembly properties of the α-helical paramyosinmolecule (Gengyo-Ando and Kagawa, ’91).

In conclusion, we report the identification of aputative phosphorylatable peptide in the para-myosin of Mytilus galloprovincialis ABRM. It is pro-posed that by observing possible effects of thesynthetic peptide or its specific antibody on thephysiology of the skinned ABRM fiber, clues to elu-cidate the functional significance of this paramyosinin muscle regulation may become apparent.

ACKNOWLEDGMENTSWe express our sincere appreciation to Drs. D.J.

Hartshorne, M.J. Siegman, and T.M. Butler, fortheir encouragement during the course of thisstudy. Thanks are also due to Dr. S. SriKanthafor his critical reading of the manuscript.

LITERATURE CITEDAchazi RK. 1979. Phosphorylation of molluscan paramyosin.

Pflugers Arch 379:197–201.Achazi RK, Dolling B, Haakshorst R. 1974. 5-HT-induzierte

Erschaffung und cyclisches AMP bei einem glatten Mollu-skenmuskel. Pflugers Arch 349:19–27.

Baguet F. 1968. The catch-state in glycerol extracted fibersfrom a lamellibranch smooth muscle (ABRM). Eur J Physiol340:19–34.

Baguet F, Gillis JM. 1968. Energy cost of tonic contraction inlamellibranch catch muscle. J Physiol 198:127–143.

Bailey K. 1956. The proteins of adductor muscle. Pubbl StnZool Napoli 29:96–108.

Castellani L, Cohen C. 1987a. Myosin rod phosphorylation andthe catch state of molluscan muscles. Science 235:334–337.

Castellani L, Cohen C. 1987b. Rod phosphorylation favorsfolding in a catch muscle myosin. Proc Nat Acad Sci USA,84:4058–4062.

Castellani L, Cohen C. 1992. A calcineurin-like phosphataseis required for catch contraction. FEBS Lett 309:321–326.

Chen G-X, Tan R-Y, Gong Z-X, Huang Y-P, Wang S-Z, Cao T-G. 1988. Paramyosin and the catch mechanism. BiophysChem 29:147–153.

Chomczynski P, Sacchi N. 1987. Single-step method of RNAisolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Anal Biochem 162:156–159.

Church GM, Gilbert W. 1984. Genomic sequencing. Proc NatAcad Sci USA 81:1991–1995.

Cohen C. 1982. Matching molecules in the catch mechanism.Proc Nat Acad Sci USA 79:3176–3178.

Cohen C, Szent-Gyorgyi AG, Kendrick-Jones J. 1971.Paramyosin and the filaments of molluscan “catch” mus-cles: I. Paramyosin: structure and assembly. J Mol Biol56:233–238.

Cohen C, Lanar DE, Parry DAD. 1987. Amino acid sequenceand structural repeats in Schistosome paramyosin matchthose of myosin. Biosci Rep 7:11–16.

Cohen C, Parry DD. 1998. A conserved C-terminal assemblyregion in paramyosin and myosin rods. J Struct Biol122:180–187.

34 S. WATABE ET AL.

Cole RA, Twarog BM. 1972. Relaxation of catch in a mollus-can smooth muscle - 1. Effects of drugs which act on theadenyl cyclase system. Comp Biochem Physiol 43A:321–330.

Collier JR. 1983. The molecular weight of ribosomal ribonucleicacids among the Protostomia. Biol Bull 164:428–432.

Cooley LB, Johnson WH, Krause S. 1979. Phosphorylation ofparamyosin and its possible role in the catch mechanism. JBiol Chem 254:2195–2198.

Cowgill RW. 1972. Susceptibility of paramyosin to proteoly-sis and its relationship to regions of different stability. Bio-chemistry 11:4532–4539.

Cowgill RW. 1974. Location and properties of sulfhydrylgroups on the muscle protein paramyosin from Mercenariamercenaria. Biochemistry 13:2467–2474.

Cowgill RW. 1975a. Proteolysis of paramyosin from Mer-cenaria mercenaria and properties of its most stable seg-ment. Biochemistry 14:503–509.

Cowgill RW. 1975b. Segments of paramyosin formed by cleav-age at sites of cysteine residues. Biochemistry 14:4277–4279.

Dahmen A, Gallin M, Schumacher M, Erttmann KD. 1993.Molecular cloning and pre-mRNA maturation of Onchocercavolvulus paramyosin. Mol Biochem Parasitol 57:335–338.

Deitiker PR, Epstein HF. 1993. Thick filament substructuresin Caenorhabditis elegans: evidence for two populations ofparamyosin. J Cell Biol 123:303–311.

Edwards HH, Johnson WH, Merrick JP. 1977. Comparison ofsolubility properties of α-paramyosin, β-paramyosin, andacid-extracted paramyosin. Biochemistry 16:2255–2260.

Elfvin M, Levine RJC, Dewey MM. 1976. Paramyosin in in-vertebrate muscles: I. Identification and localization. J CellBiol 71:261–272.

Epstein HF, Aronow BJ, Harris HE. 1976. Myosin-paramyosincofilaments: enzymatic interactions with F-actin. Proc NatAcad Sci USA 73:3015–3019.

Gengyo-Ando K, Kagawa K. 1991. Single charge on the heli-cal surface of the paramyosin rod dramatically disruptsthick filament assembly in Caenorhabditis elegans. J MolBiol 219:429–441.

Hauck R, Achazi RK. 1987. The ultrastructure of a mollus-can catch muscle during a contraction catch relaxation cycle.Eur J Cell Biol 45:30–35.

Inoue K, Waite JH, Matsuoka M, Odo S, Harayama S. 1995.Interspecific variations in adhesive protein sequences ofMytilus edulis, M. galloprovincialis, and M. trossulus. BiolBull 189:370–375.

Johnson WH. 1962. Tonic mechanisms in smooth muscles.Physiol. Rev 42(Suppl. 5):113–143.

Johnson WH, Kahn JS, Szent-Gyorgyi AG. 1959. Paramyosinand contraction of “catch muscles.” Science 130:160–161.

Kagawa H, Gengyo K, McLachlan AD, Brenner S, Karn J.1989. Paramyosin gene (unc-15) of Caenorhabditis elegans.Molecular cloning, nucleotide sequence and models for thickfilament structure. J Mol Biol 207:311–333.

Kominz DR, Saad F, Laki K. 1958. Chemical characteriza-tions of annelid, mollusc and arthropod tropomyosins. In:Committee of Muscle Chemistry of Japan, editor. Confer-ence on the chemistry of muscular contraction. Tokyo: Igaku-Shoin. p 66–76.

Laclette JP, Landa A, Arcos L, Willms K, Davis AE, Shoe-maker CB. 1991. Paramyosin in the Schistosoma mansoni(Trematoda) homologue of antigen B from Taenia solium(Cestoda). Mol Biochem Parasitol 44:287–296.

Laemmli UK. 1970. Cleavage of structural proteins duringthe assembly of the head of bacteriophage T4. Nature227:680–685.

Landa A, Laclette JP, Nicholson-Weller A, Shoemaker CB.1993. cDNA cloning and recombinant expression of collagen-binding and complement inhibitor activity of Taenia soliumparamyosin (AgB). Mol Biochem Parasitol 60:343–348.

Levine RJC, Elfvin M, Dewey MM, Walcott B. 1976. Para-myosin in invertebrate muscles - II. Content in relation tostructure and function. J Cell Biol 71:273–279.

Limberger RK, McReynolds LA. 1990. Filarial paramyosin:cDNA sequences from Dirofilaria immitis and Onchocercavolvulus. Mol Biochem Parasitol 38:271–280.

Lowy J, Millman BM, Hanson J. 1964. Structure and func-tion in smooth tonic muscles of lammellibranch molluscs.Proc Roy Soc London B, 160:525–536.

Marchand-Dumont G, Baguet F. 1975. The control mecha-nisms of relaxation in molluscan catch muscle (ABRM).Pflugers Arch 354:87–100.

Maroto M, Arredondo JJ, Roman MS, Marco R, Cervera M.1995. Analysis of the paramyosin/miniparamyosin gene. JBiol Chem 270:4375–4382.

Matsudaira P. 1987. Sequence from picomole quantities ofproteins electroblotted onto polyvinylidene difluoride mem-branes. J Biol Chem 262:10035–10038.

Matsuno A, Kanda M, Okuda M. 1996. Ultrastructural stud-ies on paramyosin core filaments from native thick filamentsin catch muscles. Tissue Cell 28:501–505.

McLachlan AD, Karn J. 1982. Periodic charge distributionsin the myosin rod amino acid sequence match cross-bridgespacings in muscle. Nature 299:226–231.

Merrick JP, Johnson WH. 1977. Solubility properties of α-reduced paramyosin. Biochemistry, 16:2260–2264.

Muhlschlegel F, Sygulla L, Frosch P, Massetti P, Frosch M.1993. Paramyosin of Echinococcus granulosus: cDNA se-quence and characterization of a tegumental antigen.Parasitol Res 79:660–666.

Nara T, Matsumoto N, Jaanecharut T, Matsuda H, YamamotoK, Irimura T, Nakamura K, Aikawa M, Oswald I, Sher A,Kita K, Kojima S. 1994. Demonstration of the target mol-ecule of a protective IgE antibody in secretory glands of Schis-tosoma japonicum larvae. Internat Immunol 6:963–971.

Nonomura Y. 1974. Fine structure of the thick filament inmolluscan catch muscle. J Mol Biol 88:445–455.

Offer G. 1990. Skip residues correlate with bends in the myo-sin tail. J Mol Biol 216:213–218.

Oleander J. 1971. Substructure of the paramyosin molecule.Biochemistry 10:601–609.

Radlick L, Johnson WH. 1982. Characterization of a para-myosin phosphokinase. Biophys J 37:36a.

Parry DAD. 1981. Structure of rabbit skeletal myosin. Analy-sis of the amino acid sequence of two fragments from therod region. J Mol Biol 153:459–464.

Pfitzer G, Ruegg JC. 1982. Molluscan catch muscle: regulationand mechanics in living and skinned anterior byssus retrac-tor muscle of Mytilus edulis. J Comp Physiol 147:137–142.

Saiki RK, Gelfand DH, Staffel S, Scharf SJ, Higuchi R, HornGT, Mullis KB, Erlich HA. 1988. Primer-directed enzymaticamplification of DNA with thermostable DNA polymerase.Science 239:487–491.

Sambrook J, Fritsch EF, Maniatis T. 1989. In: Molecular clon-ing: a laboratory manual, 2nd edition. New York.

Schriefer LA, Warterson RH. 1989. Phosphorylation of theN-terminal region of Caenorhabditis elegans paramyosin. JMol Biol 207:455–457.

Siegman MJ, Moores SU, Li C, Narayan S, Trinkle-MulcahyL, Watabe S, Hartshorne DJ, Butler TM. 1997. Phosphory-

MYTILUS PARAMYOSIN 35

lation of a high molecular weight (~600 kDa) protein regu-lates catch in invertebrate smooth muscle. J Muscle ResCell Motil 18:655–670.

Siegman MJ, Funabara D, Kinoshita S, Watabe S, HartshorneDJ, Butler TM. 1998. Phosphorylation of a twitchin-relatedprotein controls catch and calcium sensitivity of force pro-duction in invertebrate smooth muscle. Proc Nat Acad SciUSA 95:5383–5388.

Sohma H., Yazawa M, Morita F. 1985. Phosphorylationof regulatory light chain-a (RLC-a) in smooth musclemyosin of scallop, Patinopecten yessoensis. J Biochem98:569–572.

Sohma H, Morita F. 1986. Purification of a protein kinasephosphorylating myosin regulatory light chain a (RLC-a)from smooth muscle of scallop, Patinopecten yessoensis. JBiochem 100:1155–1164.

SriKantha S, Watabe S, Hashimoto K. 1990. Immunologicalcomparison between the paramyosins from adductor muscleof bivalves and from foot muscle of a gastropod. CompBiochem Physiol 96B:89–92.

Stafford III WF, Yphantis DA. 1972. Existence and inhibitionof hydrolytic enzymes attacking paramyosin in myofibril-lar extracts of Mercenaria mercenaria. Biochem Biophys ResCommun 49:848–854.

Szent-Gyorgyi AG, Cohen C, Kendrick-Jones J. 1971. Para-myosin and the filaments of molluscan “catch” muscle. II.Native filaments: isolation and characterization. J Mol Biol56:239–258.

Twarog BM. 1976. Aspect of smooth muscle function in mol-luscan catch muscle. Physiol Rev 56:829–838.

Twarog BM, Muneoka Y, Ledgere M. 1977. Serotonin anddopamine as neurotransmitters in Mytilus: block of seroto-nin receptors by an organic mercurial. J Pharmac Exp Ther201:350–356.

Vinos J, Doming A, Marco R, Cervera M. 1991. Identificationand characterization of Drosophila melanogaster para-myosin. J Mol Biol 220:687–700.

Vinos J, Maroto M, Garesse R, Marco R, Cervera M. 1992.Drosophila melanogaster paramyosin: developmental pat-tern, mapping and properties deduced from its completecoding sequence. Mol Gen Genet 231:385–394.

Watabe S, Tsuchiya T, Hartshorne DJ. 1989. Phosphoryla-tion of paramyosin. Comp Biochem Physiol 94B:813–821.

Watabe S, Hartshorne DJ. 1990. Paramyosin and the catchmechanism. Comp Biochem Physiol 96B:639–646.

Watabe S, SriKantha S, Hashimoto K, Kagawa H. 1990. Phos-phorylation and immunological cross-reactivity of para-myosin: a comparative study. Comp Biochem Physiol96B:81–88.

Winkelman L. 1976. Comparative studies of paramyosin.Comp Biochem Physiol 55B:391–397.

Woods EF, Pont MJ. Characterization of some invertebratetropomyosins. Biochemistry 10:270–276.

Yeung AT, Cowgill RW. 1976. Structural difference betweenα-paramyosin and β-paramyosin of Mercenaria mercenaria.Biochemistry 15:4654–4659.