We obtained live specimens of the Japanese pearl oyster, Pinctada fucata that were cultured in Ago Bay, Mie Prefecture, Japan. The adductor muscle, gill, mantle, and foot were dissected from each oyster body, immediately frozen in liquid nitrogen, and stored at −80˚C until use.

Total RNA was extracted from the phasic part of the adductor muscle using a conventional method [20] . First strand cDNA was synthesized using the 3’-Full RACE Core Set (TaKaRa-Bio, Ohtsu, Japan) using the total RNA as the template. Three 3’ RACE primers were designed based on the partial sequence of Pifuc-TnC: 5’-GTAGAGGACTTAAGGTGGAT-3’ for the first PCR, 5’-TAAAATCGTTAGGTGATGAT-3’ for the nested PCR, and 5’-TTCACCATAAAGGTCACCCT-3’ for the second nested PCR [18] . PCR was carried out using SapphireAmp Fast PCR Master Mix (TaKaRa-Bio) with the forward primers detailed above and the Oligo dT-3 sites adaptor primer. PCR conditions were as follows: 30 cycles of denaturation at 98˚C for 5 s, annealing at 55˚C for 5 s, and elongation at 72˚C for 20 s. The amplified DNA fragment was sequenced after insertion into a pTAC-1 vector (BioDynamics Laboratory Inc., Tokyo, Japan). 5’ RACE was carried out using the 5’ RACE system for Rapid Amplification of cDNA Ends, version 2.0 (Invitrogen, Carlsbad, CA, USA). Three primers were designed using the sequence determined by 3’ RACE: 5’-CGTCACTCCATTCTTTGAGT-3’ for synthesizing cDNA, 5’-CTCATCGTCAACTTGAAGTC-3’ for the first PCR, and 5’-AAACTTTTAAAAACTTTTTC-3’ for the second PCR. PCR was carried out using SapphireAmp Fast PCR Master Mix with the forward primers detailed above and the primers included in the kit. PCR conditions were as follows: 30 cycles of denaturation at 98˚C for 5 s, annealing at 55˚C for 5 s, and elongation at 72˚C for 10 s. The amplified DNA fragment was sequenced after insertion into a pTAC-1 vector. The determined sequence has been registered in DDBJ/EMBL/GenBank (accession number LC381286).

The primary structure of Pifuc-TnC was deduced from the nucleotide sequence determined by cDNA cloning and compared with those from mollusks using ClustalW: akazara scallop Chlamys nipponensis akazara (BAA12908), asari clam Ruditapes philippinarum (AFB83400), and squid Todarodes pacificus (Q9BLG0); arthropods: acorn barnacle Balanus nubilus (P21798), American lobster Homarus americanus (P29289), and fruit fly Drosophila melanogaster (NP_476968); nematode: Caenorhabditis elegans (BAB84566); vertebrates: chicken Gallus gallus (NP_990781), salmon Salmo salar (ACH70760), clawed frog Xenopus laevis (NP_001079408), rabbit Oryctolagus cuniculus (NP_001076114), and human Homo sapiens (NP_003270).

The 3D structure of Pifuc-TnC was predicted with SWISS-MODEL [21] using the PDB data of chicken TnC (PDB: 1YTZ) as a template. To compare the structures of different TnCs, American lobster TnC (NCBI sequence FJ790224) was also modeled following the same method.

The genome sequence including the Pifuc-TnC gene was obtained by BLAST searching the Pifuc-TnC nucleotide sequence against the P. fucata genome database [17] . The gene structure of the Pifuc-TnC gene was predicted by analyzing cDNA and genome sequences using the Splign alignment tool (NCBI) [22] .

Gene expression patterns of Pifuc-TnC in the catch and phasic muscles, gill, mantle and foot were analyzed by quantitative real-time PCR. cDNAs were synthesized using total RNA from each tissue as templates using RiverTra Ace® qPCR RT Master Mix (Toyobo Co., Ltd., Osaka, Japan). Primers and a probe were designed by Universal Probe Library Assay Design Center (Roche Diagnostics, Mannheim, Germany) using the full-length nucleotide sequence determined in this study. The primers used were: 5’-TTAACAGACGAAGAACTCGATGA-3’ (forward) and 5’-TGTCCCTGAGCCGTCTGT-3’ (reverse). Probe #94 (Roche Diagnostics) was used as a TaqMan probe. P. fucata β-actin (AF378128) was used as an internal standard. The primers used for β-actin were as follows: 5’-TCGTTCCTCGGAATGGAA-3’ (forward), 5’-TCGACATCGCATTTGAGAAT-3’ (reverse). Probe #151 (Roche Diagnostics) was used as a TaqMan probe. The PCR reaction was performed using Eagle Taq Master Mix with ROX (Roche Diagnostics).

The full-length nucleotide sequence of Pifuc-TnC was obtained through cDNA cloning and was found to contain 1775 nucleotides (nt), which includes an open reading frame of 453 nt, a 77 nt 5’ untranslated region, and a 1245 nt 3’ untranslated region (Figure 1). It encodes a protein of 150 amino acid residues in length with a predicted molecular weight of 17,400 (Figure 1).

The amino acid sequence of Pifuc-TnC was 68% homologous to both asari clam and squid TnCs and was 64% homologous to akazara scallop TnC. In contrast, the Pifuc-TnC amino acid sequence exhibited between 30 and 40% homology to TnC from arthropods, nematodes and vertebrates (Table 1).

Multiple sequence alignments of Pifuc-TnC with TnCs from other organisms revealed that Pifuc-TnC has four potential EF-hand motifs, termed sites I, II, III and IV. The Ca²⁺ binding capability of an EF-hand motif can be predicted through analysis of its amino acid sequence [11] . Site I in Pifuc-TnC is predicted to have no Ca²⁺ binding ability as the third residue, which is required to be aspartic acid, asparagine or serine in order to bind to Ca²⁺, is a lysine. In site II,

glutamic acid, phenylalanine and glutamine were identified as the third, eighth and twelfth residues, respectively, and in site III, glycine, lysine and proline were identified as the third, fifth and ninth residues, respectively. Therefore, sites II and III are also predicted to be unable to bind to Ca²⁺. Only site IV satisfied the amino acid requirements for Ca²⁺ binding, which is analogous to known TnCs from other mollusks (Figure 2 and Table 1). Therefore, out of the four potential

Ca²⁺ binding sites in Pifuc-TnC, site IV is the only EF-hand motif that is predicted to bind to Ca²⁺. To confirm our prediction, we are planning to make a recombinant Pifuc-TnC and its variants to be subjected to Ca²⁺-binding assays. A 3D model of Pifuc-TnC predicted using SWISS-MODEL was very similar to TnCs from chicken fast skeletal muscle and American lobster. The only notable difference was the presence of a short loop (four amino acids) within the α-helix connecting the site II and III EF-hand motifs (Figure 3). Similar structures were also predicted for TnC from other mollusks: akazara scallop and squid (data not shown). This structural divergence does suggest functional differences between molluscan and vertebrate TnCs. Ca²⁺ binding by vertebrate fast skeletal TnC involves all four EF-hand motifs and leads to drastic conformational changes to trigger the interaction between myosin and actin. Although it has been reported that the conformation of molluscan TnC does change upon Ca²⁺ binding, the degree of the structural change and if it consequently modifies the role of troponin in the regulation of molluscan muscle contraction remain unknown [23] . Further studies are required to unveil the role of troponin in molluscan muscle.

BLAST searching of our obtained Pifuc-TnC nucleotide sequence against the genome database of P. fucata yielded a single nucleotide sequence at scaffold 1306.1. In our previous study, we annotated a gene model (pfu_aug1.0_1306.1_22530) that was automatically predicted by the genome database to be TnC [18] . The gene model contained the predicted full-length amino acid sequence of TnC, which is identical to the sequence determined in this study. We then predicted the gene structure of Pifuc-TnC using Splign alignment of the obtained cDNA and genome sequences. Pifuc-TnC consists of five exons (Figure 4), and the start and stop codons are located in exon 1 and exon 5, respectively.

Pifuc-TnC was predominantly expressed in phasic adductor muscle, while weak expression was detected in catch adductor muscle, gill, mantle and foot (Figure 5). TnC is a key regulator in the fast contraction of vertebrate striated muscles (skeletal and cardiac muscles). The phasic adductor muscle is thought to control the quick closure of shells. Our findings suggest that TnC could be involved in the regulation of the phasic adductor muscle, rather than the catch adductor muscle. However, it has been reported that TnC was isolated from both the phasic and catch adductor muscles of the akazara scallop [24] . It is possible that the function of TnC is species-dependent.