Three calpains and ecdysone receptor in the land crab Gecarcinus lateralis: sequences, expression and effects of elevated ecdysteroid induced by eyestalk ablation

Calpains are a diverse family of intracellular Ca²⁺-dependent cysteine proteinases containing a papain-related catalytic domain(Goll et al., 2003; Mykles, 1998, 1999). In mammals, at least 14 calpain genes have been characterized(Goll et al., 2003; Suzuki et al., 2004). Calpains have also been identified in plants(Margis and Margis-Pinheiro,2003), yeast (Futai et al.,1999), fungus (Denison et al.,1995), schistosomes (Andresen et al., 1991), Hydra (GenBank #DN636661), nematode(Barnes and Hodgkin, 1996), Drosophila (Farkas et al.,2004; Jékely and Friedrich, 1999; Spadoni et al., 2003), trout (Salem et al., 2004) and lobster (Yu and Mykles, 2003). Calpains have a highly conserved cysteine protease domain (dII), which contains a catalytic triad (Cys, His, Asn) and two highly conserved Ca²⁺-binding regions required for Ca²⁺-dependent activity (Moldoveanu et al., 2002, 2004). However, in some calpains, the catalytic residues are mutated (e.g. mammalian Capn6 and Drosophila CalpC), which suggests that these calpains lack protease activity (Friedrich et al.,2004; Goll et al.,2003). The NH₂-terminal domain (dI) varies in length and sequence between different calpain genes. Autolysis of the N-terminus is associated with activation of some calpains(Goll et al., 2003). Most calpains have a C₂-like domain (dIII) that mediates Ca²⁺-dependent phospholipid binding and activation(Alexa et al., 2004; Fernandez-Montalvan et al.,2004; Friedrich et al.,2004; Hosfield et al.,1999).

Calpain genes have been organized into two general categories, based on the presence (`typical' or `EF-hand' calpains) or absence (`atypical' or`non-EF-hand' calpains) of a calmodulin-like domain (dIV) in the C-terminus(Goll et al., 2003; Huang and Wang, 2001; Suzuki et al., 2004). Domain IV contains five EF-hand motifs, the first three of which (EF-1, EF-2 and EF-3) bind Ca²⁺ with varying affinities(Hosfield et al., 1999). The fifth EF-hand (EF-5) mediates dimerization with a 28-kDa regulatory subunit in vertebrate heterodimeric calpains (Maki et al., 2002). Atypical calpains are either truncated, e.g. lobster CalpM, nematode CPL-1, mammalian nCl-2′ and DrosophilaDm-CalpA′ (Sorimachi et al.,1993; Theopold et al.,1995; Yu and Mykles,2003), or have the C-terminal region replaced by a different sequence, e.g. the `T domain' in mammalian calpain 5 (Capn5) and nematode TRA-3 (Barnes and Hodgkin,1996; Dear et al.,1997) or the `SOL domain' in mammalian SOLH and Drosophila SOL or Calpain D(Delaney et al., 1991; Friedrich et al., 2004; Kamei et al., 1998). SOL and SOLH have six Zn-finger motifs in the N-terminal region, suggesting that they are DNA-binding proteins (Delaney et al.,1991; Kamei et al.,1998).

Calpains are involved in the selective degradation of myofibrillar proteins during a molt-induced atrophy in the large claws of decapod crustaceans(Mykles, 1992, 1998). Over several weeks,there is as much as a 78% reduction in muscle mass, which facilitates withdrawal of the appendages at molt(Mykles and Skinner, 1982a; Skinner, 1966). Crustacean calpains degrade myofibrillar proteins to acid-soluble products and show about 2-fold greater activity in atrophic muscle(Mykles, 1990; Mykles and Skinner, 1982b, 1983). Lobster muscle contains four calpain activities, which were initially termed Ca²⁺-dependent proteinases (CDP I, CDP IIa, CDP IIb and CDP III; native masses 310, 125, 195 and 59 kDa,respectively; Mykles and Skinner, 1986). All require millimolar Ca²⁺ for full activity in vitro and are inhibited by cysteine protease inhibitors(Mykles and Skinner, 1983, 1986). CDP I has not been well characterized and it is the least efficient of the four calpains in degrading myofibrillar proteins (Mattson and Mykles,1993). CDP IIb is a homodimer of a 95-kDa subunit and is related to Dm-CalpA, as determined by immunological analysis(Beyette et al., 1993; Beyette and Mykles, 1997). In western blots, CDP IIa reacts with an antibody directed to a peptide sequence in the active site of mammalian calpains but not with polyclonal antibodies raised against mammalian μ- and m-calpains or Dm-CalpA(Beyette et al., 1997). These results indicate that CDP IIa differs in structure from typical calpains. CDP IIa and CDP IIb are the most effective in degrading myofibrillar proteins,such as myosin heavy and light chains, actin, tropomyosin and troponin(Mattson and Mykles, 1993). A cDNA encoding CDP III was cloned and characterized in lobster (Ha-CalpM; Yu and Mykles, 2003). Ha-CalpM lacks the calmodulin-like domain in the C-terminus that is characteristic of typical calpains. It is highly expressed in skeletal muscle, but its mRNA and protein levels do not change significantly over the molting cycle. It may be involved in restructuring and/or maintaining contractile structures in crustacean skeletal muscle (Yu and Mykles,2003).

Molt-induced muscle atrophy is coincident with increasing concentrations of ecdysteroids (ecdysone, 20-hydroxyecdysone and related compounds) in the hemolymph (Skinner, 1985). In the large claw muscles of fiddler crabs, ecdysone receptor (Up-EcR) mRNA level is increased significantly in the premolt stage(Chung et al., 1998b). (We use the term `ecdysone receptor' as per common usage; however, we recognize that it is actually an ecdysteroid receptor.) This suggests that the molting hormone, 20-hydroxyecdysone, initiates and sustains the elevated myofibrillar protein degradation that results in the reduction of muscle mass. Ecdysteroids are steroid hormones that regulate growth, development, reproduction and molting in arthropods (Chang,1993; Kozlova and Thummel,2000). Ecdysteroids regulate gene transcription after binding to a nuclear ecdysone receptor (EcR). In insects, EcR heterodimerizes with ultraspiracle (USP), a vertebrate RXR ortholog, to form a functional hormone receptor (Yao et al., 1993);it binds to a specific response element to induce expression of early response genes, such as E74 and E75(Thummel, 1996). The products of these early response genes are usually transcription factors that initiate an ecdysteroid cascade reaction (Huet et al., 1995).

Here, we report the cloning of cDNAs encoding three calpain genes(Gl-CalpM, Gl-CalpB and Gl-CalpT) and ecdysone receptor (Gl-EcR) isolated from the tropical land crab Gecarcinus lateralis. The deduced sequences of the three calpains were compared with those of calpains from other species. The expression of the three calpains in nine tissues was quantified by real-time RT-PCR. The effects of elevated ecdysteroid, induced by eyestalk ablation, on calpain and EcR mRNA levels in claw and thoracic muscles were determined. The results suggest that Gl-CalpT is involved with the initiation of myofibrillar proteolysis induced by elevated ecdysteroid.

Adult blackback land crabs, Gecarcinus lateralisFréminville, were collected from San Miguel Reserve near Fajardo,Puerto Rico. They were kept in covered plastic cages with aspen bedding moistened with tap water and were maintained at 27°C, 50% humidity and a 12 h:12 h L:D cycle. They were fed cat chow, carrots and lettuce twice a week. All eight walking legs were autotomized in some animals to induce limb regeneration and molting (Skinner,1985). The regeneration index, which is the ratio of the limb bud length to carapace width, expressed as a percentage, was used to assess limb bud growth (Skinner, 1985). Eyestalks were ablated to derepress the Y-organs to increase ecdysteroid levels in the hemolymph (Skinner,1985). Wounds were cauterized with a heated spatula to minimize loss of hemolymph.

Partial calpain cDNAs were initially obtained by nested RT-PCR using degenerate primers directed to highly conserved sequences in the protease domain of a wide variety of calpains in the GenBank database(http://www.ncbi.nlm.nih.gov),including those from nematode (GenBank accession #NP502751), fruit fly(#NP477047 and #AAD04331), rat (#NP058813) and human (#AAH08751). Conserved sequences were identified by aligning proteins using the ClustalW program(http://www.ebi.ac.uk/clustalw/index.html). Two sets of degenerate primers were designed to anneal to DNA sequences encoding LG(N/D/E)CW(L/F), EKA(Y/F)AK or (G/R)HAY(T/S)(V/I) in the catalytic domain: CPN F1, (G/A)II (C/T)T(A/G/T/C) GG(A/G/T/C) (G/A)A(A/G/T/C) TG(C/T)TGG; CPN F2, GA(G/A) AA(G/A) GC(A/G/T/C) (C/T)A(C/T) GC(A/G/T/C) AA(G/A); CPN R1, IA(C/T) (A/G/T/C)(G/C)(A/T) (G/A)TA (A/G/T/C)GC (G/A)TG (IC); CPN R2,(C/T)TT (A/G/T/C)GC (G/A)T(G/A) (A/G/T/C)GC (C/T)TT (C/T)TC. Partial sequences encoding EcR were initially obtained by nested RT-PCR using degenerate primers directed to highly conserved sequences in the DNA-binding (domain C; MMRKCQ and CRL(K/R)KC) and ligand-binding (domain E; EYALL(T/A)A and DQI(A/T)LLK)domains of EcR cDNAs in the GenBank database, as described above: EcR F1, ATG MGN MGN AAR TGY CAR; EcR F2, TGY MGN YTN VSN AAR TG); EcR R1, CNG YNA RNA RNG CRT AYT C; EcR R2, TTN ARN ARN GYD ATY TGY TC. To obtain more of the 3′sequence in the EcR ORF, a second round of nested PCR was done using two sequence-specific forward primers (cEcR F1 and cEcR F2; Table 1) and two degenerate primers to highly conserved sequences (AEIWDV and PFLAEI) in domains E and F(EcR R3, CNG YNA RNA RNG CRT AYT C; EcR R4, DAT YTC NGC NAR RAA NGG). All primers were synthesized and purified by Integrated DNA Technologies, Inc.(Des Moines, IA, USA).

Total RNA was isolated from intermolt claw muscle or primary limb regenerates (regeneration index 13–16; Skinner, 1985) using a Qiagen RNeasy mini kit (Qiagen, Inc., Valencia, CA, USA). cDNA was synthesized according to the manufacturer's protocol using the SuperScript II RNase H-reverse transcriptase first strand synthesis system (Invitrogen, Inc.,Carlsbad, CA, USA). Briefly, 12 μl of a mixture containing 1 μl oligo(dT)12–18 (500 μg ml^–1), 1 μg total RNA and 1μl 10 mmol l^–1 dNTPs was heated to 65°C for 5 min and chilled on ice for 1 min. 5× First-Strand Buffer (4 μl), 2 μl 0.1 mol l^–1 DTT and 1 μl RNaseOUT recombinant ribonuclease inhibitor (40 units μl^–1; Invitrogen, Carlsbad, CA, USA)were added, and the mixture was incubated at 42°C for 2 min. The reaction was initiated by the addition of 1 μl (200 units) of SuperScript II and incubated at 42°C for 50 min. The reaction was inactivated by heating at 70°C for 15 min.

PCR was performed using an ABI 9600 thermal cycler (Perkin-Elmer, Inc.,Wellesley, MA, USA). Claw muscle cDNA was used for the EcR PCR, and limb regenerate cDNA was used for the calpain PCR. The reactions contained 3 μl cDNA, 3 μl 10× Takara EX Taq buffer (Takara, Inc., Madison, WI, USA),2 μl 250 μmol l^–1 dNTPs, 1 μl forward primer (either CPN F1, EcR F1 or cEcR F1), 1 μl reverse primer (CPN R1, EcR R1 or EcR R3),0.2 μl Takara EX Taq DNA polymerase (5 units μl^–1) and 18.8 μl PCR-grade deionized water. Initial denaturation (95°C for 5 min) was followed by 35 amplifying cycles (either 95°C for 30 s, 55°C for 30 s and 72°C for 30 s for calpain or 95°C for 30 s, 53°C for 30 s and 72°C for 1 min for EcR) and final extension at 72°C for 7 min. For the second PCR reaction, 0.2 μl of the first PCR reaction was used, in conjunction with one of four primer pairs (CPN F1/CPN R2, CPN F2/CPN R1, EcR F2/EcR R2 or cEcR F2/EcR R4). Other reaction components and PCR conditions were the same as those in the first reaction.

The PCR products were separated by 1.2% agarose gel electrophoresis and stained with ethidium bromide. The PCR products were purified from the gel slices using the Qiaquick Gel Extraction kit (Qiagen, Inc.), ligated into TA cloning vector with the TOPO TA Cloning kit (Invitrogen, Inc.) and transformed into One Shot TOP 10 E. coli strain (Invitrogen, Inc.). Transformants were first selected by blue–white colony selection on LB agar plates containing 50 μg ml^–1 ampicillin (Sigma-Aldrich, Inc., St Louis, MO, USA) and subjected to PCR with T7 and M13-reverse vector primers to verify sizes of inserts. Plasmids were purified using the Qiagen Spin Mini prep kit and sequenced using T7 and M13-reverse vector primers (Davis Sequencing, Inc., Davis, CA, USA). If needed, gene-specific primers were used to obtain the complete sequences of both strands.

RACE (rapid amplification of cDNA ends) was used to obtain full-length mRNA sequences. Poly (A⁺) RNA was isolated from claw muscle or limb regenerate total RNA using the Oligotex mRNA kit (Qiagen, Inc.). For the 3′ sequence, the Invitrogen 3′ RACE System was used. Briefly,first-strand cDNA synthesis reactions contained 200 ng claw muscle or limb regenerate poly (A⁺) RNA and adaptor primer(5′-GGCCACGCGTCGACTAGTACTTTTTTTTTTTTTTTTT-3′). First-round PCR on 20 ng cDNA included a universal amplification primer(5′-CUACUACUACUAGGCCACGCGTCGACTAGTAC-3′) and gene-specific forward primers (Table 1) at the following conditions: denaturation at 96°C for 5 min, 35 amplification cycles (96°C for 30 s, 60°C for 30 s and 72°C for 2 min) and extension at 72°C for 10 min. Nested PCR on 30 μl of each reaction was conducted with genespecific primers (Table 1) and abridged universal amplification primer(5′-GGCCACGCGTCGACTAGTAC-3′) using the same conditions as the first-round PCR. Gel-purified PCR products were cloned into the TA vector and sequenced as described above.

SMART™ RACE cDNA Amplification kit (BD Biosciences, Inc., Palo Alto,CA, USA) was used to obtain the 5′ sequences of each gene. The first-strand cDNA synthesis reaction contained 3 μl poly(A+) RNA (50 ng for calpain primers and 100 ng for EcR primers), 1 μl 10 mmol l^–1 5′ CDS primer [5′-(T)25N-1N-3′] and 1μl 10 mmol l^–1 SMART II A oligo(5′-AAGCAGTGGTATCAACGCAGAGTACGCGGG-3′) and was incubated at 68°C for 2 min. After chilling the reaction for 2 min on ice, 2 μl 5× First-Strand buffer (250 mmol l^–1 Tris-HCl, pH 8.3;375 mmol l^–1 KCl; 30 mmol l^–1MgCl₂), 1 μl 20 mmol l^–1 DTT, 1 μl 10 mmol l^–1 dNTPs and 1 μl PowerScript reverse transcriptase (BD Biosciences Clontech, Mountain View, CA, USA) were added. The reaction was covered with 20 μl paraffin oil and incubated at 42°C for 1.5 h in an ABI 9600 thermal cycler. The reaction mixture was diluted 10-fold with autoclaved distilled water and used for first-round PCR with 10×Universal Primer A Mix (0.4 mmol l^–15′-CTAATACGACTCACTATAGGGCAAGCAGTGGTATCAACGCAGAGT-3′ and 2 mmol l^–1 5′-CTAATACGACTCACTATAGGGC-3′) and gene-specific reverse primers (Table 1) under the following conditions: denaturation at 96°C for 5 min, 35 amplification cycles (either 96°C for 30 s, 65°C for 15 s and 72°C for 3 min for calpains or 96°C for 30 s, 66°C for 15 s and 72°C for 3 min for EcR) and extension at 72°C for 10 min. Second-round PCR was conducted using nested gene-specific primers(Table 1) and nested universal primer A (10 mmol l^–1,5′-AAGCAGTGGTATCAACGCAGAGT-3′). The PCR conditions were the same as those used for first-round PCR. Gel-purified PCR products were cloned into the TA vector and sequenced as described above. Continuous sequences of three calpains were obtained by PCR using primer pairs to the start and stop codons to verify the sequence of each ORF.

Phylogenetic relationships between calpain sequences were determined with Treeview, which is a program that displays ClustalW result files(http://taxonomy.zoology.gla.ac.uk/rod/treeview/treeview_manual.html).

Total RNA was isolated from tissues using the RNeasy mini kit according to the manufacturer's instructions (Qiagen, Inc.). Total RNA (1 μg) was DNase-treated (Invitrogen, Inc.) and reverse-transcribed using the SuperScript II RNase H-reverse transcriptase first strand synthesis system (Invitrogen,Inc.). End-point PCR reactions were performed in an ABI 9600 thermal cycler using TaKaRa Ex Taq HotStart polymerase (Takara, Inc.), gene-specific primers(Table 2) and either 2 μl(calpains) or 5 μl (EcR) of the first strand cDNA mixture as template. The PCR conditions for calpains were an initial denaturation at 95°C for 4 min, 35 polymerization cycles (denaturation at 94°C for 30 s, annealing at 60°C for 30 s and extension at 72°C for 30 s), and final extension at 72°C for 2 min. PCR reactions were analyzed by separating all of the 20μl reaction volume on 2% agarose ethidium bromide-stained gels. Primers were designed using IDT BioTools program(http://biotools.idtdna.com/gateway).

For quantitative RT-PCR, transcript levels of the three calpains (Gl-CalpB,Gl-CalpM and Gl-CalpT) and ecdysone receptor (Gl-EcR) were determined by real-time PCR using a Cepheid Smart Cycler instrument (Cepheid, Sunnyvale, CA,USA) and sequence-specific primers (Medler and Mykles, 2003; Yu and Mykles, 2003). Transcript levels of the four genes were normalized to the mRNA level of elongation factor 2 (EF2), which was constitutively expressed. The PCR products were ligated into TOPO 2.1 vector using TOPO TA Cloning kit (Invitrogen, Inc.). Standard curves were generated using serial dilutions (10 fg to 10 ng) of plasmids containing either Gl-CalM, Gl-CalpB,Gl-CalpT or Gl-EF2 inserts (Medler and Mykles, 2003; Yu and Mykles,2003). Reaction mixtures contained 2 μl LightCycler FastStart Reaction Mix (Roche, Indianapolis, IN, USA; 10× buffer, Fast Taq DNA polymerase and dNTPs), 2 μl 25 mmol l^–1 MgCl_2,2 μl cDNA template, 1 μl forward primer(Table 3; 10 pmolμl^–1), 1 μl reverse primer(Table 3; 10 pmolμl^–1) and 12 μl PCR-grade water. The PCR conditions were denaturation at 96°C for 5 min, 40 amplification cycles (96°C for 20 s, 65°C for 15 s and 72°C for 30 s), and melting curve detection(60°C+0.2 deg. s^–1). PCR products were evaluated by melting temperature analysis and separation on 2% agarose ethidium bromide-stained gels (Medler and Mykles,2003; Yu and Mykles,2003).

The Statview 5.0.1 Program (SAS Institute, Inc., Cary, NC, USA) was used for statistical analyses. One-way analysis of variance (ANOVA) was used to compare the copy numbers determined by real-time PCR. Values were log-transformed due to high levels of variability between samples and to correct for the correlation between mean and variance(Medler and Mykles, 2003). Pair-wise post-ANOVA comparisons used a Bonferroni test with an experiment-wise error rate of 0.05.

Nested PCR with degenerate primers generated products of the expected size from cDNA synthesized from primary limb regenerate mRNA. Our rationale was that since the limb regenerate contained various tissues (e.g. muscle,epidermis, nerve, connective tissue, hemocytes, blood vessels), a greater number of calpain species would be obtained than by using a single tissue. After cloning into a plasmid vector, 10 different clones were selected for sequencing. The calpain-like sequences encoding a part of the protease domain fell into four categories: the first was similar to lobster CalpM (accession#AY124009); the second was similar to Drosophila CalpA (accession#NP477047) and CalpB (accession #NP524016); the third was similar to human Capn5 (accession #CAA71584) and C. elegans TRA-3 (accession#NP502751); and the fourth was similar to C. elegans clp-1 (accession#NP741237). 3′ and 5′ RACE using nested sequence-specific primers yielded full-length sequences of three of the four initial sequences,designated Gl-CalpB, Gl-CalpM and Gl-CalpT (G. lateralis calpains B,M and T). Expression of each calpain in limb regenerates was confirmed by RT-PCR. However, the clp-1-like sequence was not amplified and therefore not characterized further; it may have resulted from contamination of the limb regenerate mRNA with mRNA from a parasitic nematode that occurs in some individuals. The full ORFs were confirmed by PCR using specific primer sets containing the start codon and stop codon sequences for each calpain.

The characteristics of the cDNAs encoding the three land crab calpains are summarized in Table 4. The domain organization of the land crab calpains is compared with that of calpains from lobster (Ha-CalpM), Drosophila (Dm-CalpA and Dm-CalpB)and nematode (TRA-3) in Fig. 1. The DNA and deduced amino acid sequences of Gl-CalpB, Gl-CalpM and Gl-CalpT are presented in Figs 2, 3 and 4, respectively. The deduced amino acid sequence of Gl-CalpB was 61% identical to Dm-CalpB, 51% identical to Dm-CalpA and 50% identical to human Capn3 (accession #NP-775110.1). The deduced amino acid sequence of Gl-CalpM was most similar to that of Ha-CalpM(66% identity) and Dm-CalpA (48% identity). The amino acid sequence of Gl-CalpT was 47% identical to human Capn5 (accession #O15484) and 41%identical to C. elegans TRA-3 (accession #AAB60256).

Gl-CalpB had the four-domain organization of typical calpains: N-terminal(dI), catalytic (dII), C₂-like (dIII) and calmodulin-like (dIV)(Figs 1, 2). Gl-CalpM had the conserved catalytic and C₂-like domains but lacked the C-terminal calmodulin-like domain (Figs 1, 3). A minor, alternatively spliced variant of Gl-CalpM (Gl-CalpM′) was obtained with 3′ RACE;the deduced polypeptide is truncated 958 bp from the start codon, which results in the absence of the catalytic asparagine in domain II(Fig. 3). Gl-CalpT resembled other calpains from domains I to III, but dIV was replaced with a T domain found in Capn5 and TRA-3 (Figs 1, 4).

Multiple amino acid sequence alignment showed high similarity in domains II and III in calpains from nematode (TRA-3), arthropods (Gl-CalpM, Gl-CalpB,Gl-CalpT, Dm-CalpA and Dm-CalpB) and mammals (calpains 1 and 3)(Fig. 5). All have the conserved catalytic triad (C, H, N) and two non-EF hand Ca²⁺-binding sites in dII. A C₂-like sequence containing an acidic loop in dIII was common to all calpains. The major difference between the two CalpM sequences was that Ha-CalpM had two acidic amino acid insertions, one (DDSDD) near the end of dII and the other (DDDDDDDDDDRG) in the C₂ acidic loop region, that were absent in Gl-CalpM. Unlike a muscle-specific mammalian Capn3, calpains from arthropods and nematode lacked an insertion sequence in dII. Domain IV, when present, was well conserved from arthropods to mammals, although Dm-CalpA had an insertion sequence between EF-1 and EF-2 not found in any other calpain(Fig. 6). The sequence of the T domain of Gl-CalpT was similar to that of other `T domain' calpains(Fig. 7).

To study the sequence relationships of the three land crab calpains, a phylogenetic analysis was done based on the deduced amino acid sequences of dII and dIII of calpains from arthropods, nematode and mammals(Fig. 8). The sequences clustered in four groups. TRA-3, Capn5 and Gl-CalpT were clustered as a distinct group, even though the T domain sequences were not included in the analysis. Mammalian Capn1 and Capn3 were grouped together and were more closely related to calpains with an EF-hand domain. Crustacean-specific Ha-CalpM and Gl-CalpM were grouped separately from arthropod A/B-type calpains(Gl-CalpB and Dm-CalpA and B).

The initial 700-bp product from nested PCR using degenerate EcR primers was 95% identical to the deduced amino acid sequence of the EcR from fiddler crab(accession #AAC33432). 3′ RACE using sequence-specific primers failed. If the 3′ UTR in the Gl-EcR is as long as it is in fiddler crab EcR(∼4 kb), it may be difficult to amplify with 3′ RACE. Additional 3′ sequence was obtained by PCR using gene-specific forward primers(cEcR F1 and cEcR F2; Table 1)and degenerate primer sets (EcR F3 and EcR F4; see Materials and methods). 5′ RACE yielded a 300-bp product, which extended the 5′ sequence another 100 bp. The consensus partial sequence of Gl-EcR was 1005 bp and encoded a deduced 335-amino acid sequence encompassing part of the C domain and all of the D and E domains (Fig. 9). The partial amino acid sequence of Gl-EcR was aligned with full-length sequences of EcR-encoding genes from fiddler crab (Up-EcR) and locust (Lm-EcR) (Fig. 10). The Gl-EcR sequence was 93% identical with homologous regions of Up-EcR and 66%identical with Lm-EcR.

End-point RT-PCR showed that most tissues expressed the three calpains(Fig. 11). All tissues expressed Gl-CalpB. The molting gland (Y-organ) did not express Gl-CalpM or Gl-CalpT. Eyestalk ganglia did not express Gl-CalpT. The mRNA levels in nine tissues were quantified by real-time PCR(Fig. 12). Three general expression patterns were observed. In skeletal muscle(Fig. 12A,B), Gl-CalpM and Gl-CalpB were expressed at significantly higher levels than Gl-CalpT; there was no significant difference between Gl-CalpM and Gl-CalpB. In heart, gill,thoracic ganglion, digestive gland and testis, Gl-CalpB was expressed at higher levels, often significantly, than Gl-CalpM and Gl-CalpT; there was no significant difference between Gl-CalpM and Gl-CalpT(Fig. 12C–G). In ovary and integument, Gl-CalpM was expressed at higher levels than Gl-CalpB or Gl-CalpT, although the difference in the means was only significant in ovary;there was no significant difference between Gl-CalpB and Gl-CalpT in either tissue (Fig. 12H,I). Interestingly, the striated muscles showed different patterns. In skeletal muscle, Gl-CalpM and Gl-CalpB were expressed about 100-fold and 20-fold higher than Gl-CalpT in claw and thoracic muscles, respectively(Fig. 12, compare A and B with C). By contrast, Gl-CalpB was expressed in heart about 5-fold higher than Gl-CalpM and about 2.6-fold higher than Gl-CalpT. Gonadal tissues also differed in expression pattern. Gl-CalpM was expressed in ovary about 4.4-fold higher than Gl-CalpB and about 10.5-fold higher than Gl-CalpT, while Gl-CalpB was expressed in testis about 100-fold higher than Gl-CalpM and about 32-fold higher than Gl-CalpT (Fig. 12,compare G with H).

As claw muscle atrophy coincides with elevated ecdysteroid levels during premolt, the effects of ecdysteroids, induced by eyestalk ablation, on the expression of the EcR and the three calpains were determined in claw and thoracic muscles. The thoracic muscle served as a non-induced control, as it does not undergo premolt atrophy (Moffett,1987). EcR mRNA was used as an indicator of tissue sensitivity to molt induction, as it is upregulated in response to elevated ecdysteroids in insects (Sun et al.,2003).

The X-organ/sinus gland complex in the eyestalk is the primary source of molt-inhibiting hormone (MIH), a neuropeptide that inhibits ecdysteroid synthesis and secretion by the Y-organs. Eyestalk ablation, therefore, is used as an effective strategy to rapidly stimulate the Y-organs and chronically increase the level of ecdysteroids in the hemolymph(Skinner, 1985). Ecdysteroid levels were significantly higher one day after eyestalk ablation(80.8±22.3 ng ml^–1, mean ± 1 S.D., N=4) compared with intact controls (31.9±5.9 ng ml^–1, N=5; P=0.002). Ecdysteroid levels in 3-day eyestalk-ablated animals were not determined, although previous work showed that ablation results in chronically elevated ecdysteroid levels(McCarthy and Skinner,1977).

The EcR, calpain and EF2 mRNA levels in thoracic and claw muscles were quantified by real-time PCR. EF2, which is constitutively expressed, was used as an internal standard to normalize the PCR reactions. In thoracic muscle,eyestalk ablation had no significant effect on expression of the Gl-EcR and the three calpains (Fig. 13B). By contrast, Gl-EcR expression in claw muscle increased about 15-fold one day after eyestalk ablation but then decreased to about 2.8-fold above the level in intact animals three days after eyestalk ablation(Fig. 13A). Gl-CalpT mRNA paralleled the expression of Gl-EcR in claw muscle; it increased about 19.3-fold above the level in intact animals one day after eyestalk ablation and about 4.3-fold higher three days after eyestalk ablation. Gl-CalpM and Gl-CalpB expression was not affected by eyestalk ablation. The relationship between the three calpain mRNAs and the EcR mRNA is presented in Fig. 14. The expression of Gl-EcR and CalpT was highly correlated in both claw muscle and thoracic muscle from intact and eyestalk-ablated animals(Fig. 14, bottom panels). There was no correlation between EcR expression and either Gl-CalpM or Gl-CalpB in either muscle (Fig. 14, top and middle panels, respectively).

cDNAs encoding three full-length calpains were obtained from land crab limb regenerates by PCR cloning. Four calpain genes [Dm-CalpA, B, C and D (or SOL)] occur in the Drosophila genome(Friedrich et al., 2004; Goll et al., 2003). Sequence alignments indicate that the arthropod calpains fall into five categories; a proposed classification is presented in Table 5. Surprisingly, only members of the A/B type occur in both insects and crustaceans. Gl-CalpB,Dm-CalpA and Dm-CalpB appear to be the arthropod form of typical calpains, as they have the four-domain organization of mammalian calpains 1, 2 and 3 (Figs 2, 6; Goll et al., 2003; Jékely and Friedrich,1999; Suzuki et al.,2004). Dm-CalpC was initially identified by the genome project as CG3692 and was predicted to have nine transmembrane domains in the N-terminal region. However, a recent study showed that Dm-CalpC actually has a much shorter N-terminal domain that lacks the membrane-spanning segments; it is catalytically inactive because all three catalytic amino acid residues are mutated (Spadoni et al.,2003). Members of the D (SOL), M and T types are atypical calpains, as they lack dIV. Gl-CalpM and Ha-CalpM are truncated at the C-terminal end of dIII (Figs 1, 3), while Gl-CalpT has the T domain in place of dIV in the C-terminus (Figs 1, 4). D calpains have six zinc-finger motifs in the N-terminal `SOL' domain in place of dIII and dIV in the C-terminal region (Delaney et al.,1991).

Gl-CalpB appears to be the only typical calpain in crustaceans, as it is the only land crab calpain with the C-terminal EF-hand domain IV(Fig. 2). Sequence alignment analysis of domains II and III and its estimated mass (∼89 kDa) indicate that Gl-CalpB is most closely related to Drosophila calpains A and B(Farkas et al., 2004; Jékely and Friedrich,1999; Pinter et al.,1992). Gl-CalpB combines features found in Dm-CalpA or Dm-CalpB. It has no insertion sequence in dIV, like Dm-CalpB, but has a dI length similar to that in Dm-CalpA (Fig. 1; Emori and Saigo,1994; Jékely and Friedrich, 1999; Theopold et al., 1995). These data suggest that Dm-CalpA and Dm-CalpB are derived from duplication of a common ancestral gene and that no such duplication occurred in crustaceans.

The M-type calpains are unique to crustaceans. Gl-CalpM from land crab has the highest structural similarity with Ha-CalpM, the first calpain cloned from a crustacean (Yu and Mykles,2003). Both encode polypeptides of about 66 kDa that lack dIV. Unlike other truncated forms, crustacean CalpM does not appear to be produced by alternate splicing of a typical calpain gene. Dm-CalpA′, for example,is an alternatively spliced transcript of Dm-CalpA(Theopold et al., 1995). In mammals, nCL-2′, which lacks domains III and IV, is produced by alternative splicing of the nCL-2 (Capn8) gene(Sorimachi et al., 1993). We were not successful in obtaining longer cDNAs containing the dIV region using 3′ RACE PCR with various sequence-specific primer sets. The absence of a longer mRNA is supported by immunoblot results using an antibody raised against a 28-amino acid N-terminal sequence of Ha-CalpM; only two proteins with masses of 62 and 68 kDa were detected in lobster muscles(Yu and Mykles, 2003). A truncated, alternatively spliced isoform of Gl-CalpM, Gl-CalpM′, was isolated by 3′ RACE. Gl-CalpM′ is likely to be catalytically inactive, as the alternative transcript has a stop codon inserted before the Asn residue of the catalytic triad in dII. As it is expressed at very low levels, it was not characterized further.

The Gl-CalpT is a member of the T domain calpain family. Its amino acid sequence has the highest sequence identity with human Capn5 and nematode TRA-3(Figs 5, 7). TRA-3 was first isolated as a sex determinant of the soma and germ line in hermaphrodites of C. elegans (Barnes and Hodgkin,1996). TRA-2A is a substrate for TRA-3, and cleavage of TRA-2A by TRA-3 generates a peptide that has feminizing activity(Sokol and Kuwabara, 2000). Two mammalian calpains, Capn5 and Capn6, have the same structure as Gl-CalpT and TRA-3. However, Capn6 is catalytically inactive, as one or two residues of the catalytic triad (C, H, N) are mutated in human (K, H, N) and mouse (K, Y,N) (Dear et al., 1997; Matena et al., 1998). Interestingly, the Drosophila genome apparently lacks a CalpT-like gene (Friedrich et al., 2004; Goll et al., 2003).

It is likely that all three crustacean calpains have Ca²⁺-dependent proteinase activity. The deduced amino acid sequences of the protease domain are highly conserved, including the three amino acid residues essential for catalytic activity(Fig. 5). In addition, the three calpains contain two well-conserved non-EF hand Ca²⁺-binding sites in dII and a C₂-like Ca²⁺/phospholipid-binding site in dIII (Fig. 5), both of which are important in Ca²⁺-dependent activation(Alexa et al., 2004; Moldoveanu et al., 2002, 2004). In the absence of Ca²⁺, the catalytic residues are misaligned and the substrate binding cleft is disrupted; binding of Ca²⁺ to domains II and III activates the enzyme by driving the realignment of the active site residues(Moldoveanu et al., 2002, 2004; Strobl et al., 2000). The interaction of residues R104 and E333 in dII of mammalian calpain provides cooperativity between the two Ca²⁺-binding sites(Moldoveanu et al., 2004);these two residues are conserved in the crustacean genes. Moreover, calpains lacking the calmodulin-like domain retain Ca²⁺-dependent activity(Hata et al., 2001; Sokol and Kuwabara, 2000).

Four Ca²⁺-dependent proteinase (CDP) activities have been characterized biochemically in crustacean muscle(Mykles and Skinner, 1986). These activities differ in native mass and subunit composition(Table 6). Ha-CalpM appears to encode CDP III, based on similar masses and chromatography properties(Yu and Mykles, 2003); the native enzyme consists of a single polypeptide(Table 6). Gl-CalpB may encode CDP IIb, as the subunit mass (95 kDa) of the purified lobster CDP IIb(Beyette et al., 1993; Beyette and Mykles, 1997) is similar to the estimated mass (89 kDa) of the deduced Gl-CalpB amino acid sequence. In addition, CDP IIb and the Dm-CalpA 95-kDa gene product share immunological properties. Polyclonal antibodies raised against lobster CDP IIb and Dm-CalpA protein cross-react, while a polyclonal antibody raised against a conserved 20-amino acid sequence around the cysteine residue in the active site of mammalian μ- and m-calpains (GATRTDICQGALGDCWLLAA) does not react with either Gl-CalpB or Dm-CalpA (Beyette et al., 1997). Analysis of this same sequence in Gl-CalpB(GATRFDVKQGELGDCWLLAA) indicates that four residues (Thr, Ile, Cys and Ala) in mammalian calpains are replaced by Phe, Val, Lys and Glu in Gl-CalpB(Fig. 5). The replacement of two uncharged residues with two charged residues may explain why the antibody does not react with lobster CDP IIb. The Gl-CalpB polypeptide probably does not associate with a regulatory subunit, as the sequence in the putative EF-5 is the least conserved with vertebrate heterodimeric calpains (e.g. Capn1; see Fig. 6). This is consistent with the homodimeric structure of the native CDP IIb(Beyette and Mykles, 1997).

The identity of Gl-CalpT with CDP I or IIa is less certain. A polyclonal antibody raised against the 20-residue mammalian active site sequence reacts with a 60-kDa protein in immunoblots of a partially purified preparation of lobster CDP IIa (Beyette et al.,1997). However, eight of the 20 residues around the catalytic cysteine differ between Capn1 and Gl-CalpT(Fig. 5), making it less likely that the antibody would recognize the CalpT sequence. Furthermore, the deduced amino acid sequence of Gl-CalpT has an estimated mass of ∼74 kDa, which is significantly greater than the 60 kDa estimated from immunoblots. CDP I is not well characterized. Its large native mass (310 kDa) suggests it is a multimer of the proteins encoded by CalpB, CalpT and/or an additional unidentified calpain gene. CalpM can be excluded as a component of CDP I, as the Ha-CalpM antibody does not react with any proteins in CDP I fractions eluting from a gel filtration column (Yu and Mykles,2003). Further work is required to reconcile the CalpB and CalpT cDNAs with the calpain activities.

Three general patterns of calpain expression are observed in the nine tissues from adult intermolt land crabs analyzed by real-time PCR. The first pattern, in which both Gl-CalpB and Gl-CalpM are expressed at higher levels than Gl-CalpT, is found in skeletal muscles(Fig. 12A,B). The second pattern, in which Gl-CalpB is dominant, is found in heart, gill, thoracic ganglion, digestive gland and testis (Fig. 12C–G). The third pattern, in which Gl-CalpM mRNA level is highest, is found in ovary and integument(Fig. 12H,I). Differences in calpain expression are also observed in insect. Although Drosophilacalpains are expressed throughout development, Dm-CalpA, Dm-CalpB and Dm-CalpC differ in their mRNA levels and tissue distribution. In early embryos,Dm-CalpA is localized in the anterior pole and posterior surface regions(Emori and Saigo, 1994), while Dm-CalpB has a uniform distribution(Farkas et al., 2004). During nuclear division, Dm-CalpA is more evenly distributed and becomes associated with the precleavage furrows before cellularization(Emori and Saigo, 1994). In late embryos, Dm-CalpB is preferentially expressed in the trachea and foregut(Farkas et al., 2004). In larvae, Dm-CalpB mRNA and protein levels decline, whereas Dm-CalpC mRNA is moderately elevated (Farkas et al.,2004; Spadoni et al.,2003). Both Dm-CalpB and Dm-CalpC are expressed highly in the salivary glands of third instar larvae(Farkas et al., 2004; Spadoni et al., 2003). In adults, Dm-CalpA mRNA is present in many tissues, such as ovary, brain,ventral ganglion, midgut and heart, but not in the indirect flight muscles(Amano et al., 1997; Theopold et al., 1995). Dm-CalpB is also expressed in ovary(Farkas et al., 2004). The tissue distribution of Dm-CalpC in adults has not been reported. Such dynamic changes in localization and distribution suggest that the different calpains carry out specialized functions in arthropod tissues.

Ha-CalpM was first identified as a muscle-specific calpain, as it is expressed at the highest levels in lobster skeletal muscles(Yu and Mykles, 2003). The analysis of Gl-CalpM, however, shows that it is expressed in ovary and integument at levels higher than that in skeletal muscle. It might first appear that the tissue expression pattern of Gl-CalpM deviates more from that of Ha-CalpM. However, the apparent differences are much less, when one considers the fiber-type compositions of the muscles. The levels of the Ha-CalpM mRNA are about 4-fold higher in fast muscle (cutter claw closer and deep abdominal flexor muscles) than in slow muscle (crusher claw closer)(Yu and Mykles, 2003). The claw closer and thoracic muscles in the land crab are composed of only slow fibers (Mykles, 1988; Mykles and Skinner, 1981). If one excludes the fast muscles, the ratios of Gl-CalpM expression in slow muscle in relationship with that in most of the other tissues are similar to those of Ha-CalpM. An exception is the difference in CalpM expression in the integument. In lobster, Ha-CalpM mRNA level in the integument is about 13%that in crusher claw closer muscle (Yu and Mykles, 2003), whereas Gl-CalpM mRNA level is about 18-fold and 43-fold greater in integument than that in the thoracic muscle and claw muscle, respectively (Fig. 12,compare A, B and I). Gl-CalpM was also highly expressed in ovary. Unfortunately, the lobster ovary was not analyzed for Ha-CalpM expression(Yu and Mykles, 2003).

A cDNA encoding a partial sequence of EcR was obtained from land crab claw muscle mRNA. The domain organization of Gl-EcR is similar to that of fiddler crab EcR (Chung et al., 1998a; Durica et al., 2002), as well as other nuclear steroid receptors (Renaud and Moras, 2000). The N-terminal A/B domain is the least conserved in amino acid sequence and length (Cherbas et al., 2003; Hu et al.,2003; Onate et al.,1998). In insects, receptors differing in A/B domains are generated by alternate transcriptional start sites or alternative splicing(Schubiger et al., 2003; Segraves and Woldin, 1993; Talbot et al., 1993). By contrast, the fiddler crab EcR has one A/B domain isoform but has several alternative spliced isoforms around domain E(Durica et al., 2002). The C domain is highly conserved among different nuclear steroid receptors and primarily serves as a DNA-binding domain (DBD) containing two zinc-finger motifs. The E domain is a moderately conserved region of about 250 amino acids that serves as the ligand-binding domain (LBD)(Billas et al., 2003; Grebe et al., 2003; Wang et al., 2000). Domain E of many nuclear streroid receptors can also interact with numerous proteins,such as homo/heterodimeric partners, co-repressors and co-activators(Hu et al., 2003; Shibata et al., 1997). Both the DBD and LBD of Gl-EcR and Up-EcR are closely related to those domains in insect EcR (Fig. 10) and therefore probably have the same function. The D domain is a hinge region connecting the DBD with the LBD and is important in ligand binding(Grebe et al., 2003).

Gl-EcR and Gl-CalpT expression are upregulated by eyestalk ablation in claw muscle but not in thoracic muscle. Premolt atrophy occurs specifically in the claw muscle (Moffett, 1987; Mykles and Skinner, 1982a),suggesting that sensitivity to ecdysteroid may be important in initiating muscle protein degradation. The expression of Gl-EcR and Gl-CalpT was highly correlated in claw and thoracic muscles in intact and eyestalk-ablated animals(Figs 13, 14). These results suggest that expression of Gl-EcR and Gl-CalpT is linked. One possibility is that the EcR/RXR complex binds directly to the promoter of the Gl-CalpT gene,which induces its expression. Another possibility is that ecdysone early response genes, such as E75 or E74 are induced by the EcR/RXR complex, and they, in turn, induce the expression of Gl-CalpT. In Manduca sexta, the EcR/RXR complex induces E75 within 30 min (Zhou et al.,1998). The lower expression of Gl-EcR and Gl-CalpT in claw muscle three days after eyestalk ablation (Fig. 13A) suggests a feedback inhibition in response to sustained elevated ecdysteroids that is not mediated by the neurosecretory center in the eyestalk. In insects, 20-hydroxyecdysone inhibits ecdysteroid production in the molting gland (Beydon and Lafont,1983; Sakurai and Williams,1989), which is associated with changes in the expression and phosphorylation state of certain USP isoforms(Song and Gilbert, 1998). In crustaceans, ecdysteroid inhibits ecdysteroidogenesis in the Y-organ, but the mechanism is not known (Dell et al.,1999).

The role of each calpain in premolt claw muscle atrophy remains to be established. Unlike mammalian skeletal muscle, in which calpains are restricted to initial disassembly of sarcomeric elements(Goll et al., 2003; Jackman and Kandarian, 2004),crustacean calpains can carry out both disassembly and subsequent degradation of myofibrillar proteins (Mykles,1998). Only Gl-CalpT is upregulated in response to elevated ecdysteroids, suggesting it is involved in initiating the atrophy program through the proteolytic modification of signaling proteins. This is analogous to the modification of TRA-2A by TRA-3 in mediating feminization of C. elegans embryos (Sokol and Kuwabara,2000). In mammals, limited proteolysis of protein kinases and phosphatases by calpain often alters their biochemical properties(Goll et al., 2003; Mykles, 1998). The ubiquitous tissue expression and lack of ecdysteroid regulation of Gl-CalpB suggests that it has a housekeeping function. There is a preferential degradation of thin filaments during atrophy (Ismail and Mykles, 1992; Mykles and Skinner, 1981). CDP I and CDP IIb more efficiently degrade thin filaments than CDP IIa and CDP III(Mattson and Mykles, 1993). CDP III is encoded by CalpM (Yu and Mykles, 2003). Gl-CalpB probably encodes CDP IIb(Table 6), but Gl-CalpT cannot be identified with either CDP I or CDP IIa without further analysis. Neither Ha-CalpM mRNA nor protein levels in lobster claw muscle change during the molting cycle (Yu and Mykles,2003). If CalpB and CalpM are involved, their activities are not controlled at the transcriptional level. Calpains are regulated post-translationally by phosphorylation and/or endogenous activators or inhibitors (Friedrich, 2004; Goll et al., 2003; Mykles, 1998). We propose that an increase in the relative activity of Gl-CalpB (CDP IIb) compared with the other calpains mediates the preferential degradation of thin filaments during premolt.

This research was supported by a grant from the National Science Foundation(IBN-9904528). We thank Hector C. Horta (Puerto Rico Department of Natural Resources) and Hector J. Horta for collecting land crabs; Sharon Chang for ecdysteroid radioimmunoassays; Laura Baker, Emma James, Lindsey Progen,Stephanie Spenny and Kristin Van Ort for animal care; and Drs Scott Medler and Kara Lee for critical reading of the manuscript.