Mechanisms of protein degradation in mantle muscle and proposed gillremodeling in starved Sepia officinalis

Simon G. Lamarre,1,2 Delphine Ditlecadet,2,3 David J. McKenzie,2,4 Laure Bonnaud,5,6

and William R. Driedzic2,3

1Department of Biochemistry, Memorial University of Newfoundland, St. John’s, Newfoundland, Canada; 2StationMéditerranéenne de l’Environnement Littoral, Navals, Sète, France; 3Ocean Sciences Centre, Memorial University ofNewfoundland, St. John’s, Newfoundland and Labrador, Canada; 4UMR 5119 Ecologie des Systèmes Marins Côtiers,Université Montpellier II, Montpellier cedex 5, France; and 5Museum National d’Histoire Naturelle (MNHN), DépartementMilieux et Peuplements Aquatiques, UMR Biologie des Organismes et Ecosystemes Aquatiques, MNHN, Centre National de laRecherche Scientifique 7208, Institut de Recherche pour le Developpement 207, Université Pierre et Marie Curie, Paris,France; and 6Université Paris Diderot, Sorbonne Paris Cité, Paris, France

Submitted 16 February 2012; accepted in final form 24 May 2012

Lamarre SG, Ditlecadet D, McKenzie DJ, Bonnaud L, DriedzicWR. Mechanisms of protein degradation in mantle muscle and pro-posed gill remodeling in starved Sepia officinalis. Am J Physiol RegulIntegr Comp Physiol 303: R427–R437, 2012. First published May 30,2012; doi:10.1152/ajpregu.00077.2012.—Cephalopods have relatively highrates of protein synthesis compared to rates of protein degradation,along with minimal carbohydrate and lipid reserves. During fooddeprivation on board protein is catabolized as a metabolic fuel. Theaim of the current study was to assess whether biochemical indices ofprotein synthesis and proteolytic mechanisms were altered in cuttle-fish, Sepia officinalis, starved for 7 days. In mantle muscle, fooddeprivation is associated with a decrease in protein synthesis, asindicated by a decrease in the total RNA level and dephosphorylationof key signaling molecules, such as the eukaryote binding protein,4E-BP1 (regulator of translation) and Akt. The ubiquitination-proteasome system (UPS) is activated as shown by an increase in thelevels of proteasome �-subunit mRNA, polyubiquitinated protein, andpolyubiquitin mRNA. As well, cathepsin activity levels are increased,suggesting increased proteolysis through the lysosomal pathway.Together, these mechanisms could supply amino acids as metabolicfuels. In gill, the situation is quite different. It appears that during thefirst stages of starvation, both protein synthesis and protein degrada-tion are enhanced in gill. This is based upon increased phosphoryla-tion of 4E-BP1 and enhanced levels of UPS indicators, especially 20Sproteasome activity and polyubiquitin mRNA. It is proposed that anincreased protein turnover is related to gill remodeling perhaps toretain essential hemolymph-borne compounds.

cathepsin; cephalopod; metabolic enzymes; proteasome; polyubiqui-tin; RNA; starvation; triglyceride

CEPHALOPODS TYPICALLY FEED upon protein-rich diets and aregenerally high in protein content with low reserves of lipid andcarbohydrates. Most of the lipid is present in the digestivegland with very low lipid reserves in other tissues (33). Studiestracking the incorporation of radiolabeled phenylalanine intothe protein pool show that in feeding cephalopods, proteinaccretion is favored by relatively low rates of protein degra-dation compared to rates of protein synthesis (6, 18, 27). InSouthern dumpling squid (Euprymna tasmanica), whole bodyprotein degradation increases in minimally fed animals (18).Furthermore in octopus, Octopus vulgaris, the oxygen con-

sumption-nitrogen excretion ratio increased over the first 5days of starvation to values that imply a mixed lipid andprotein oxidation, and by day 8 the ratio decreased to apredominantly protein-based metabolism (5). In cuttlefish, Se-pia officinalis, oxygen consumption is maintained for the first6 days of starvation, and thus, rates of metabolic fuel deliverymust remain constant (15). Collectively, the above implies thatduring periods of food deprivation, the limited lipid reservesare called upon first, in parallel with, then subsequently,followed by, an obligate dependence on the use of protein as ametabolic fuel. The objective of the current study was to assesswhether biochemical indices of protein breakdown and to someextent protein synthesis in various tissues are altered duringfood deprivation in S. officinalis.

Mechanisms of protein degradation have been well studiedin vertebrate animals. One pathway is the ubiquitin-proteasomesystem (UPS) that involves the ubiquitination of protein fol-lowed by degradation via proteasomes. In mammalian celllines and especially in mammalian muscle during starvation,the ubiquitin-proteasome pathway is the dominant route ofprotein degradation (24, 40). Here, we assess, for the first timein any cephalopod, the activity of 20S proteasome, levels ofpolyubiquitinated protein, transcript levels of proteasome�-subunit (a component of the catalytic subunit), and transcriptlevels of polyubiquitin. In addition, we determined the degreeof phosphorylation of the signaling molecules Akt (also knownas PKB) and the eukaryote binding protein, 4E-BP1. Akt isconsidered to play a role in numerous signaling pathways,including the UPS, whereas, 4E-BP1 is a regulator of thetranslation step of protein synthesis. Phosphorylation of Aktcan result in phosphorylation of 4E-BP1, leading to the acti-vation of protein synthesis. The dephosphorylated state of Aktand 4E-BP1 should lead to a curtailment of protein synthesis(1, 14).

Protein degradation may also occur via the transport ofprotein into lysosomes with subsequent proteolysis by lowpH-active cathepsins and by nonlysosomal intracellular calci-um-dependent calpains. Activation of the lysosomal pathwayhas been noted in white muscle of starved rainbow trout(Oncorhynchus mykiss) and Atlantic cod (Gadus morhua) (7,16). Cathepsin-like activity has also been reported in viscera ofS. officinalis (4). Here, we determine in numerous tissuescathepsin A, B, H, L-like activity at pH 5.5, cathepsin D, E-like

Address for reprint requests and other correspondence: W. R. Driedzic,Ocean Sciences Centre, Memorial Univ. of Newfoundland, St. John’s, NL,Canada, A1C 5S7 (e-mail: wdriedzic@mun.ca).

Am J Physiol Regul Integr Comp Physiol 303: R427–R437, 2012.First published May 30, 2012; doi:10.1152/ajpregu.00077.2012.

0363-6119/12 Copyright © 2012 the American Physiological Societyhttp://www.ajpregu.org R427

activity at pH 2.5, and calpain-like protease activity in mediumcontaining calcium at neutral pH. Following proteolysis, thereleased amino acids must be metabolized via aerobic metab-olism and the citric acid cycle to release energy to sustain ATPproduction. As such, the activity of enzymes involved in the finaltrafficking of amino acids aspartate aminotransferase (AspAT),alanine aminotransferase (AlAT), and glutamate dehydrogenase(GDH), as well as the mitochondrial enzyme citrate synthase (CS)were also determined.

To provide a context for protein degradation, levels of RNAin various tissues and triglyceride content in digestive glandwere assessed, as these are hallmarks of nutritional status incephalopods. This contention is based on the following obser-vations. Within individual tissues of O. vulgaris rates ofprotein accretion increased linearly with RNA/protein ratio(19). As well, whole body RNA per gram, fractional rates ofprotein synthesis, and fractional rates of protein accretion allfollowed similar regressions when plotted against total bodyweight in E. tasmanica (25). However, during starvation, theRNA/protein ratio decreased in mantle of the octopus Eledonecirrhosa (18), and the RNA/DNA ratio decreased in paralarvaeof the squid Loligo opalescens (39), as well as whole (�67 mg)juvenile S. officinalis (8). Most compelling in relation to thecurrent experiment is the finding that in mantle muscle from S.officinalis, there was a decrease in RNA/DNA ratio and RNA/gdry weight after 4 to 6 days of starvation (25). Taken together,these data imply that the level of RNA is an important deter-minant of rates of protein synthesis and when food-deprived,the RNA pool is decreased presumably as a means to decreaseprotein synthesis. Digestive gland mass decreased with starva-tion in E. cirrhosa (18) and in S. officinalis 4 days of fooddeprivation resulted in about a 35–50% decrease in its mass (7,25). In S. officinalis, digestive gland triglycerides constituteabout ½ of the lipids in fed animals and decrease duringstarvation, whereas, phospholipids that make up the bulk of theremainder of the lipid pool are mobilized to only a limitedextent. As such, digestive gland triglyceride serves as anenergy source during the first 10 days of food deprivation andis a clear indicator of the level of starvation (7).

This study presents novel information about the mechanismsof protein degradation in cephalopods under physiologicalconditions and provides an intellectual connection between theUPS and signaling processes. Our major findings are thatduring food deprivation, the UPS and cathepsin activity areincreased in mantle muscle, but in gill, there appears to be atleast a transient increase in protein turnover (both synthesis anddegradation).

MATERIALS AND METHODS

Animals. Cuttlefish (Sepia officinalis) were captured by net by alocal fisherman. Animals were transferred to the Station Méditer-ranéenne de l’Environnement Littoral (Sète, France) and maintainedin free-running sea water. Studies were conducted in May 2010(experiment 1) and May 2011 (experiment 2). For experimentalpurposes, animals were held in individual shallow tanks (200 � 45 �25 cm) with fine gravel on the bottom. Water temperatures variedbetween 16°C and 18°C in 2010 and between 20°C and 22°C in 2011.S. officinalis were either fed daily or food-deprived for a period of 7days. In experiment 1, the final body masses were 296 � 63 g and282 � 31 g for fed and starved animals, respectively. In experiment 2,the final body masses were 261 � 40 g and 364 � 127 g for fed and

starved animals, respectively. There was no statistically significantdifferences in body mass among the four experimental groups. Thelarger average mass in starved than in fed animals is presumably dueto variability in initial body masses and not weight loss during 7 daysof food deprivation. In experiment 1, fed animals were offered two seabream (Sparus aurata) per day with a body mass of about 4 g,resulting in a daily ration of about 3% body weight per day. Inexperiment 2, ration was increased in an attempt to generate a greaterrange in nutritional status. In this case, fed animals were offered 4 or5 sea bream per day, resulting in a daily ration of about 7% bodyweight per day. Following 7 days, the animals were anesthetized in3% ethanol in sea water and the ventral nerve cord was severed behindthe head. Sections of mantle muscle, retractor muscle, systemic heart,gill, digestive gland, and, in some cases, hemolymph were harvested,frozen on dry ice, and stored at �80°C for future analysis. Animalprotocols were approved by the Animal Care Committee at MemorialUniversity of Newfoundland.

Lipid analysis. Total triglyceride level in the digestive gland wasdetermined as described by Collison et al. (10). Tissue was homog-enized in ground glass [60 mg/ml chloroform:methanol (2:1)]. Thehomogenate was transferred to a centrifuge tube and allowed to standfor 2 h with frequent shaking. Thereafter, 0.2 vol of water was added;the sample was vortexed and centrifuged for 5 min at 2,000 g; and thebottom layer was transferred to a glass vessel and allowed to dry. Thelipid was resuspended in ethanol (250 mg tissue/ml), and triglycerideslevels were determined via Sigma kit TR0100 (Sigma-Aldrich, St.Louis, MO).

20S proteasome activity. In experiment 1, tissues were homoge-nized in 5 vols of ice-cold lysis buffer using a polytron (2 � 10 s) andcentrifuged at 13,000 g at 4°C for 60 min. The lysis buffer was 50 mMTris, 0.1 mM EDTA, 0.007% �-mercaptoethanol, pH 8.0. In experi-ment 2, the lysis buffer included a phosphatase inhibitor cocktail(containing cantharidin, p-bromolevamisole oxalate, and calyculin A;kit no. P0044; Sigma-Aldrich). Immediately after centrifugation, theprotein concentration of the supernatant was determined using theBradford protein assay kit (Bio-Rad, Hercules, CA) with BSA asthe standard. The fresh supernatant was used for the assay of 20Sproteasome activity.

For experiment 1, the 20S proteasome activity was measuredaccording to methods used by Shibatani and Ward (36). Fifty micro-grams of protein from the supernatant was incubated for 60 min with40 �M of the fluorogenic substrate LLVY-AMC (Biomol Interna-tional, Plymouth Meeting, PA) in 100 �l of 100 mM Tris buffer (pH8.0) containing 0.0285% SDS (this concentration was determinedexperimentally; see section below). Reactions were conducted withand without the addition of 50 �M of both MG-115 and MG-132(Enzo Life Sciences, Farmingdale, NY), two inhibitors of the 20Sproteasome. The LLVY, MG115, and MG132 were dissolved inDMSO. The final concentration of DMSO in the assay was 0.6% and1.1% in assays without and with inhibitors, respectively. The reactionwas stopped by adding 300 �l of 1% SDS and 1 ml of 0.1 mM sodiumborate (pH 9.1). The fluorescence was determined in a single-cellspectrofluorometer at excitation and emission peaks of 370/430 nm.The inhibitor-sensitive activity was considered as the 20S proteasomeactivity and reported as arbitrary fluorescence units·h�1·50 �g pro-tein�1.

The optimal concentration of SDS for the activation of the 20Sproteasome activity was empirically determined in gill and mantle oftwo S. officinalis by assaying the activity as above using SDSconcentrations ranging from 0 to 0.06%. Maximal 20S proteasomeactivation was detected at 0.0285% of SDS and rapidly decreased atconcentrations above 0.030%. The activation profile of SDS wasconsistent between mantle and gill. Accordingly, all assays wereperformed at SDS concentration of 0.0285%.

In experiment 2, the 20S proteasome activity was measured usinga fluorescence microplate reader (BioTek Synergy HT, Winooski,VT) in kinetic mode instead of an end-point as in experiment 1. Each

R428 PROTEIN DEGRADATION IN SEPIA

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00077.2012 • www.ajpregu.org

well consisted of 50 �g of supernatant protein, 100 �l of 100 mM Trisbuffer containing 40 �M LLVY-AMC and 0.0285% SDS, with orwithout MG-115 and MG-132. The increase of fluorescence waslinear for at least 30 min at 25°C and was typically recorded for 10min. The activity is reported as above.

Dot-blot analysis of polyubiquitinated proteins. Aliquots of thesupernatant for tissue homogenates used for 20S proteasome activitieswere quickly frozen and stored at �80°C until dot blot analysis of thelevel of polyubiquitinated proteins. For each sample, 50 �g of proteinof the supernatant described above were spotted on a nitrocellulosemembrane. The membrane was blocked with 5% BSA. The polyu-biquitinated proteins were detected using a specific antibody (cloneFK1; Enzo Life Sciences). Dots were revealed by enhanced chemi-luminescence (ECL) after secondary detection using a horseradishperoxidase (HRP) conjugated goat anti-mouse IgM (Abcam, Cam-bridge, MA). Dot intensity was quantified by densitometry usingImageJ (National Institutes of Health, Bethesda, MD).

Western blot analysis. Aliquots of the supernatant of tissue homog-enates used for 20S proteasome activities in experiment 2 werequickly frozen and stored at �80°C until analysis of levels of Akt,phospho-Akt (serine 473), 4E-BP1, and phospho-4E-BP1 (threonine37/46). Tissue homogenates (50 �g of protein) were subjected to SDSpolyacrylamide gel electrophoresis (SDS-PAGE), blotted on PVDFmembranes and the blots were blocked using 5% BSA. The mem-branes were probed with protein-specific antibodies obtained fromCell Signaling Technologies (Beverley, MA). The phosphorylatedform of the proteins was quantified first; then, the membranes werestripped and probed with the antibody recognizing the total protein.Secondary detection was performed using a goat anti-rabbit HRP-conjugated antibody, and bands were revealed using ECL. Bandintensity was quantified using ImageJ and the ratio of phospho-Akt/Akt and phospho-4E-BP1/4E-BP1 were calculated.

Metabolic enzymes. The fresh supernatant obtained from prepara-tions for 20S proteasome activity in experiment 1 was diluted 1:1 withhomogenization buffer and used directly for assay of metabolicenzymes. All assays for metabolic enzymes were previously used withcephalopod tissues (12, 17). Specific conditions were as follows: foraspartate aminotransferase (AspAT; EC 2.6.1.1), 50 mM imidazole,40 mM L-aspartate, 0.025 mM pyridoxal phosphate, 0.15 mM NADH,and excess malate dehydrogenase were initiated with 10 mM 2-oxo-glutarate, at pH 7.0; for alanine aminotransferase (AlAT; EC 2.6.1.2),50 mM imidazole, 200 mM L-alanine, 0.025 mM pyridoxal phosphate,0.15 mM NADH, and excess lactate dehydrogenase were initiatedwith 10 mM 2-oxoglutarate, at pH 7.0; for glutamate dehydrogenase(GDH; EC 1.4.1.3), 100 mM Tris, 200 mM NH4Cl, 2 mM ADP, and0.2 mM NADH were initiated with 5 mM 2-oxoglutarate, at pH 8.0;and for citrate synthase (CS; EC 2.3.3.1), 100 mM TRIS, 0.25 mMDTNB [5,5= dithiobis-(2-nitrobenzic acid)], and 0.4 mM acetyl CoA,initiated with 4 mM oxaloacetate, at pH 8.0.

All enzymes were assayed at 15°C in a spectrophotometer with1-ml cuvettes. Activity was expressed as �mol·min�1·mg protein�1

on the basis of millimolar extinction coefficients of 6.22 for NADH-linked reactions at 340 nm and 13.6 for CS at 412 nm.

Protease assays. In experiment 1, tissues were homogenized in 10vols of medium containing 20 mM Tris, 1 mM EDTA, 100 mM KCl,and 0.1% mercaptoethanol at pH 7.5 and centrifuged for 5 min at10,000 g. In experiment 2, the fresh supernatant obtained frompreparations for 20S proteasome activity was used. In all cases, thesupernatant was diluted to 1% (tissue mass/volume) for enzymeanalysis. Cathepsin-like activity was determined in medium at pH 5.5(97 mM citric acid, 5.8 mM Na2HPO4, 40 mM cysteine) and inmedium at pH 2.5 (43.5 mM citric acid, 112 mM Na2HPO4, 40 mMcysteine), modified from Mycek (29). Calpain-like activity was de-termined in medium containing 20 mM Tris, 1 mM EDTA, 10 mMCaCl2, 100 mM KCl, and 0.1% mercaptoethanol, at pH 7.5. As-says were based on the degradation of fluorescently labeled casein(BODIPY-FL-conjugated casein) that releases fluorescent cleavage

products (21, 38). Assays followed the manufacturer’s instructions inan EnzChek Protease Assay Kit E6638 (Molecular Probes, Eugene,OR). Briefly, 100 �l of diluted homogenate was added to a microplatewell followed by 100 �l of the enzyme assay buffer containing 10�g/ml BODIPY-casein. The fluorescence (excitation/emission peaks � 485/535 nm) was recorded immediately and after an incubation period of60 min at room temperature. The final fluorescence was subtractedfrom the initial value to determine the relative enzyme activity. Theassay was linear with respect to substrate concentration. Fluorescencewas determined with a Beckman Coulter DTX microplate reader inexperiment 1 and with a BioTek Synergy HT microplate reader inexperiment 2.

Proteasome �-subunit and polyubiquitin expression. Frozen sam-ples were weighed, and thereafter, total RNA was extracted byhomogenization in TRIzol reagent (Invitrogen, Burlington, ON, Can-ada). Extracted RNA was quantitated by UV absorption at 260 nm.Integrity and purity were verified prior to cDNA synthesis by agarosegel electrophoresis and by calculating the OD260/280 ratio, respec-tively. Ten micrograms of RNA were treated with DNase using theTURBO DNA-free kit (Ambion, Austin, TX). One microgram ofDNase-treated RNA was reverse transcribed using random hexamersand Moloney murine leukemia virus reverse transcriptase (Invitrogen,Carlsbad, CA). cDNA so generated was used for real-time quantita-tive PCR (qPCR). DNase treatment and reverse transcription wereperformed following manufacturer’s instructions.

Expression of proteasome �-subunit and of polyubiquitin wasquantified as levels of their respective mRNA. qPCR was done usingprepared cDNA with normalization to 16S rRNA (AB240155.1) usingSYBR Green I dye chemistry and the 7300 real-time PCR system(Applied Biosystems, Foster City, CA) or the Stratagene Mx3000Psystem (Stratagene, La Jolla, CA) in experiments 1 and 2, respec-tively. Primers used are listed in Table 1 and were designed fromsequences specific to S. officinalis for the genes considered. Se-quences of the two genes were obtained from an expressed sequencetags (EST) library (Genoscope CEA project) obtained from S. offici-nalis embryos cDNA and identified following automatic annotations(3). Relevancy of annotations was confirmed by blasting relevantknown sequences against EST, tc, and singlets databases throughSAMS blast interface. One tc was automatically annotated as ubiqui-tin (30 ESTs), five tcs were annotated as proteasome subunit (5, 4, 7,3, 44 ESTs). For proteasome, tc was chosen for its closest homologyto proteasome � subunit. [proteasome (prosome, macropain) subunit,� type, 4 (PSMB4)]. For all transcripts, efficiency and specificity ofthe reactions were confirmed before running the samples. Efficienciesof the reactions (%) were estimated for both qPCR systems and in alltissues using a five-point serial dilution, and all ranged between 90and 110%. Specificity of each reaction was verified by running a meltcurve analysis and by running the reaction products on a 2% agarosegel. Both methods showed a single specific product of the expectedsize for all genes. Samples were then run in duplicate in separatereactions. qPCR reactions were performed in a total volume of 25 �lcontaining 12.5 �l power SYBR Green PCR Master Mix (AppliedBiosystems); 50 nM of each forward and reverse primers and 2.5 ng

Table 1. Amplicon sizes and sequences of primers used forproteasome �-subunit, polyubiquitin, and 16S rRNAreal-time quantitative PCR analysis

Transcript Direction_Sequence (5=�/�3=)AmpliconSize, bp

Proteasome (�-subunit) F_ATAATGCTGGCTGCCGACACT 102R_TGCGGCAATCACAGTAGCATC

Polyubiquitin F_CAACTCTCCTACTGGACGAAAGC 70R_TGTCCAGAGCGCTAAACAACT

16S rRNA F_AATTTCTCCGGTCCTTTCGTR_TGCAGTAGCTTTAGGTGGTGAA 125

R429PROTEIN DEGRADATION IN SEPIA

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00077.2012 • www.ajpregu.org

of cDNA for proteasome �-subunit and polyubiquitin or 0.2 ng cDNAfor 16S rRNA. The qPCR program consisted of 1 cycle at 50°C for 2min, 1 cycle at 95°C for 10 min, and 40 cycles of (95°C for 15 s and60°C for 1 min). The cycle threshold was determined using the 7300PCR Detection System SDS software version 1.2.3 from AppliedBiosystems (experiment 1) or the MxPro QPCR v4.00 software fromStratagene (experiment 2). The relative quantity of proteasome �-sub-unit and of polyubiquitin mRNA was determined following the Pfafflmethod (32). This mathematical model takes into account the respec-tive efficiencies of the gene considered.

Statistical analyses. Values are expressed as means � SE. Signif-icant differences involving two conditions were assessed with atwo-tailed Student’s t-test, and in the case of multiple conditions, aone-way ANOVA with Tukey’s post hoc test. Data that are presentedas ratios were log transformed prior to statistical analysis. In all casesP � 0.05 was considered to be significant.

RESULTS

Nutritional status. There was no difference in hemolymphglucose content between fed (1.6 � 0.1 mM; n � 4) andstarved (1.7 � 0.1 mM; n � 5) S. officinalis.

Starvation was associated with a decrease in digestive glandtriglyceride content in both experiments 1 and 2 (Fig. 1, A andB). The starved S. officinalis in experiment 1 showed the lowestlevel of digestive gland triglyceride (2.1 � 2.1 mg/g tissue) andis considered to be the group that should have placed thehighest demands upon protein as a metabolic fuel. The fed S.officinalis in experiment 2 had the highest level of triglyceridein the digestive gland (64.7 � 10.7 mg/g tissue) and isconsidered to be the best nourished of the four groups. Therewas no significant correlation between hepatopancreas triglyc-eride content and body mass within each of the four groups norwhen all of the fed or all of the starved animals were consid-ered together.

In experiment 1, food deprivation was associated with asignificant decrease in RNA in heart (Fig. 1C). In experiment2, there was a significant decrease in RNA in mantle and atendency for a decrease in retractor muscle (P � 0.075)

between fed and starved animals (Fig. 1D). Fig. 2 presents therelationship between triglyceride level in digestive gland, as anindex of the nutritional state, and tissue RNA content, takinginto consideration data from both experiments. Triglyceridelevel in the digestive gland was linearly related to RNA inmantle (y � 0.15x � 22; R2 � 0.38; P � 0.02) and retractormuscle (y � 0.08x � 7; R2 � 0.36; P � 0.03). As such, lowlevels of triglyceride were associated with lower levels ofRNA. There was no obvious correlation between triglyceridelevel in digestive gland and RNA content in heart. Althoughthe data are limited, it appears that the situation in gill is morecomplex. When data from fed and starved animals were ana-lyzed separately, the starved group showed a linear correlationbetween the two variables (y � 0.01x � 15; R2 � 0.99; P �0.003). The triglyceride level in digestive gland and gill RNAlevel in the fed group also showed a tendency to correlate butwith a much steeper slope (y � 0.028x �12; R2 � 0.77; P �0.11).

Overall, the changes in RNA content induced by starvationwere small on an absolute level compared with tissue differ-ences in RNA. For instance, in experiment 2, the significantlydifferent level of RNA in mantle muscle between fed andstarved S. officinalis was 126 �g/g, whereas, the differencebetween gill and mantle RNA level was �2,500 �g/g. Acomparison, including values from both fed and starved indi-viduals, provides a generalized picture of the tissue distributionof RNA (Fig. 3). RNA level in gill is significantly higher thanin all other tissues, while RNA level in heart is higher than inmantle and retractor muscle.

Components of the ubiquitin-proteasome system. In experi-ment 1, components of the UPS were elevated in mantle ofstarved compared with fed S. officinalis (Fig. 4, left). Morespecifically, there were significant increases of the transcriptlevels of the proteasome �-subunit and polyubiquitin. Also, theamount of polyubiquitinated protein was higher in starvedanimals. The increase in mantle polyubiquitinated protein wasmost substantial as amounts were below detectable levels in

Fig. 1. Triglyceride content of digestive glandand tissue RNA levels in fed and starved S.officinalis. Left: experiment 1. Right: experi-ment 2. A and B: triglyceride levels. C andD: RNA levels. Solid bars represent fed ani-mals, while open bars represent starved ani-mals. *Significant difference between fed andstarved animals. Values are expressed asmeans � SE. n sizes are as follows. Experi-ment 1: triglyceride, 3; RNA: gill (fed), 2; gill(starved), 3; all other groups, 6–8. Experi-ment 2: triglyceride (fed), 5; triglyceride(starved), 4; RNA: gill (fed), 3; gill (starved),4; all other fed groups, 3; all other starvedgroups, 4.

R430 PROTEIN DEGRADATION IN SEPIA

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00077.2012 • www.ajpregu.org

two of three samples in the fed animals. Mean values of polyu-biquitinated protein were also about 2-fold higher in gill and heartalthough these were not statistically significant, perhaps becauseof the semiquantitative nature of the assay. All other measureswere similar between fed and starved S. officinalis.

In experiment 2, changes of components of the UPS wereevident in gill (Fig. 4, right). There were significant increasesin starved vs. fed animals in proteasome activity and polyu-biquitin transcript level. The amount of polyubiquitinated pro-tein also tended to be higher (P � 0.063), and the mean levelof proteasome �-subunit transcript was higher. All other com-parisons in heart, mantle, and retractor were similar betweenfed and starved S. officinalis.

Given that total RNA is considered to be one determinant ofprotein synthesis and 20S proteasome activity represents oneaspect of protein degradation, the relationship between thesetwo parameters was assessed. In both experiments 1 and 2,

when data for all tissues from both fed and starved animalswere taken into consideration, there was a linear correlationbetween 20S proteasome activity and RNA level (experiment1: y � 0.004x 10; R2 � 0.41; P � 0.0002 ) (experiment 2:y � 0.58x 1,290; R2 � 0.77; P � 0.0001) (Fig. 5).

Phosphorylation state of signaling molecules. The phos-phorylation state of the signaling molecules 4E-BP1 and Aktwas determined in experiment 2 (Fig. 6). The ratio of phos-phorylated 4E-BP1/4EBP1 total protein significantly decreasedin mantle but significantly increased in gill following fooddeprivation. Following starvation, there was a significant de-crease in the ratio of phosphorylated Akt/Akt total protein inmantle and retractor muscle. There was no change in thephosphorylated Akt/Akt total protein in gill.

Protease activities. Overall, there were only minimalchanges in the maximal activity of the three commonly studiedproteases (Fig. 7). In experiment 1, starvation resulted in asignificant increase in cathepsin-like activity at pH 5.5 in mantle.There was also a tendency for cathepsin-like activity at pH 2.5 toincrease in mantle (P � 0.085). In experiment 2, there was asignificant increase in calpain activity in retractor. There were nosignificant differences in any other of the comparisons.

Metabolic enzymes. There were no significant differences inactivities of AspAT, AlAT, GDH, or CS between fed andstarved animals in any tissue (Fig. 8). Heart displayed the mostvigorous activity of all in the enzymes measured. Furtheranalysis revealed a linear correlation between 20S proteasomeactivity and CS activity in heart (y � 0.02x 10; R2 � 0.64;P � 0.002) (Fig. 9).

DISCUSSION

Biochemical and physiological processes are often com-pared between cephalopods and fishes, two distantly related

Fig. 2. Triglyceride content of digestive glandvs. tissue RNA levels in fed and starved S.officinalis. A: mantle. B: retractor. C: heart.D: gill. Solid symbols represent fed animals,while open symbols represent starved animals.Squares show levels from experiment 1. Circlesshow levels from experiment 2.

Fig. 3. RNA level in various tissues of S. officinalis. Both fed and starvedanimals from experiments 1 and 2 are included in the analysis. Values areexpressed as means � SE. a,b,cValues with different letters indicate a statisti-cally significant difference. n size for gill is 12, while for all others, it is 21.

R431PROTEIN DEGRADATION IN SEPIA

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00077.2012 • www.ajpregu.org

groups that are competitors in the same ecosystems (11, 26,30, 31). Such comparisons provide fundamental insightsinto conserved features that favor species performance. Inthe following discussion on the impact of starvation in S.officinalis, reference is made to studies with fishes wherepossible.

Digestive gland triglycerides and tissue RNA levels arerelated. In keeping with previous work, food deprivation wasassociated with a decrease in triglyceride content in digestivegland of S. officinalis (7). This finding is important in settingthe context of the current study. The experimental design ofdifferent feeding levels, possibly coupled with the difference inwater temperature, nutritional status at the time of capture,and/or holding times/feeding regimes prior to food deprivation,resulted in a range of triglyceride contents that provides aplatform for comparisons. It is unlikely that scaling effectshave a major impact on the variables in this study given that

there was no statistically significant difference in body massamong the experimental groups, and triglyceride content in thehepatopancreas was not correlated with body mass. Moreover,on the basis of the scaling relationships presented in Grigoriouand Richardson (15), the metabolic rate of the group with thelowest mass would be within 28% of the group with the highestmass. The strength of the information is that there were twoextreme groups of fed vs. starved animals, regardless of howthat end point was achieved. The experiment 1 data bestillustrate what occurs once triglyceride resources are essen-tially depleted. Digestive gland triglyceride level also de-creased in experiment 2; however, there were still reserves atthe time of sampling. As such, experiment 2 is considered torepresent an earlier stage of the starvation response. The datacollectively reveal variable digestive gland triglyceride levelsthat reflect a range of nutritional states and that are useful inrelating to other variables.

Fig. 4. Components of the ubiquitin-protea-some system in various tissues of fed andstarved S. officinalis. Left: experiment 1. Right:experiment 2. A and B: 20S proteasome activity(fluorescent units·min�1·50 �g protein�1). Cand D: proteosome �-subunit mRNA levels(relative amount). E and F: polyubiquitinatedprotein (relative amount). G and H: polyubiq-uitin mRNA levels (relative amount). Solidbars represent fed animals, while open barsrepresent starved animals. *Significant differ-ence between fed and starved animals. Valuesare expressed as means � SE. n sizes are asfollows: 20S proteasome activity-Experiment1: gill (fed), 2; gill (starved), 3; heart (fed), 5;heart (starved), 7; all other conditions, 4. Ex-periment 2: all conditions, 4. Proteasome�-subunit mRNA–Experiment 1: gill (fed), 2;gill (starved), 3; all other conditions, 5. Exper-iment 2, gill (fed), 3; gill (starved), 4; heart(fed), 3; heart (starved), 4; mantle (fed andstarved), 4; retractor (fed), 3; retractor(starved), 4. Polyubiquitinated protein–Experi-ment 1, gill (fed), 2, gill (starved), 3; heart (fed),2; heart (starved), 6; mantle (fed), 3; mantle(starved), 4; retractor (fed and starved), 6. Ex-periment 2: mantle (starved), 3; all other con-ditions, 4. Poylubiquitinated mRNA–Experi-ment 1, mantle (fed and starved), 5. Experiment2, gill (fed) 3; gill (starved), 4; heart (fed), 3;heart (starved), 4; mantle (fed and starved),4; retractor (fed), 3; retractor (starved), 4.Note: it is not possible to directly comparemeasurements between experiments due tothe use of different instrumentation andslightly different techniques.

R432 PROTEIN DEGRADATION IN SEPIA

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00077.2012 • www.ajpregu.org

Total RNA content was highest in S. officinalis gill, inter-mediate in heart, and substantially lower in mantle and retrac-tor muscle when data from all groups were considered to-gether. The content pattern is similar to that observed in O.vulgaris and fishes (e.g., Atlantic cod and cunner Tautogo-labrus adspersus). Further to this point, RNA content corre-lates with rates of protein synthesis on a tissue-specific basis inthese species (13, 20, 22). It is probable that under routinefeeding conditions, total RNA content is a determinant ofprotein synthesis in S. officinalis, as well, and that rates ofprotein synthesis rank order in the same manner in S. officinalistissues as that of other species.

The impact of food deprivation on tissue RNA content notedhere was consistent with earlier studies in S. officinalis. (25)and other cephalopods (18, 39). Levels decreased in mantlemuscle during the first stages of starvation and in heart duringmore severe food deprivation. More revealing and unique tothis study is the relationship between digestive gland triglyc-eride level and tissue RNA content. These variables weresignificantly correlated in mantle and retractor muscles withthe slope being much steeper in mantle. There was no corre-lation between digestive gland triglyceride level and heartRNA content. These findings indicate that the RNA content inmantle is more sensitive to decreases in overall nutritionalstatus than either retractor muscle or heart, as indicated bydigestive gland triglyceride. Response at the level of total RNA

content could contribute to a decrease in protein synthesis inthe absence of food.

The situation in gill is much more complex. The averagelevel of total RNA was higher in starved than in fed animals.Although the data are limited, total RNA content very clearlydecreased in association with decreases in digestive glandtriglyceride level in starved S. officinalis. In fed animals thereis a suggestion for such a relationship with a tentatively greaterslope than in starved animals. This implies that in a comparisonof any two animals with the same level of triglyceride in thedigestive gland, a starved specimen will have higher levels ofRNA in the gill than a S. officinalis that is feeding. It appearsthat totally food-deprived animals better defend gill proteinsynthesis as nutritional status falls. The implications of this arediscussed below.

Ubiquitin-proteasome system is activated. This paper pro-vides the first measurements of elements of the UPS in anycephalopod. All tissues displayed 20S proteasome activity,polyubiquinated protein, and transcripts for proteasome �-sub-unit (i.e., the catalytic subunit), and polyubiquitin. In rainbowtrout, 20S proteasome activity and total ubiquitin protein arehigher in gill than in numerous other tissues (23), which isqualitatively similar to the current information with S. offici-nalis. These findings suggest that gill tissue has the highesttissue-specific rate of protein degradation via the UPS in bothcephalopods and fish. When all groups and tissues were con-sidered together, there was a correlation between 20S protea-some activity and RNA content in both experiments, suggest-ing that the two variables are both linked to protein turnover.This is consistent with total RNA being an index of protein

Fig. 5. 20S proteasome activity vs. RNA level. A: experiment 1. B: experiment2. Values for both fed and starved animals for all tissues (gill, heart, mantle,and retractor) sampled were considered in the analysis. Both regressions aresignificantly different from zero. Note: it is not possible to directly comparemeasurements between experiments due to the use of different instrumentationand slightly different techniques.

Fig. 6. Levels of signaling molecules in various tissues of fed and starved S.officinalis. A: ratio of phosphorylated 4E-BP1/4E-BP1 total protein. B: ratio ofphosphorylated Akt/Akt total protein. Solid symbols represent fed animals,while open symbols represent starved animals. *Significant difference betweenfed and starved animals. Values are expressed as means � SE. n sizes for gilland mantle, 4; n sizes for retractor, 3.

R433PROTEIN DEGRADATION IN SEPIA

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00077.2012 • www.ajpregu.org

synthesis and the UPS being mostly responsible for the deg-radation of damaged proteins.

In experiment 1, where the nutritional status of the S.officinalis is considered to be the lowest, there were significantincreases in mantle polyubiquinated protein, and levels ofproteasome �-subunit and polyubiquitin mRNA. There was no

change in 20S proteasome activity but, given the increase inproteasome �-subunit transcript, a longer period of starvationmay have led to such. This might also indicate that regulationof the UPS degradation pathway is not primarily at the level of20S proteasome content and activity. Regardless, significantchanges in three components of the UPS in mantle strongly

Fig. 7. Maximal activity levels of proteasesin various tissues of fed and starved S. offi-cinalis. Left: experiment 1. Right: experiment2. A and B: cathepsin-like activity at pH 5.5.C and D: cathepsin-like activity at pH 2.5. Eand F: calpain-like activity. Data are ex-pressed as flourescent units·min�1·mg pro-tein�1. Solid symbols represent fed animals;open symbols represent starved animals.*Significant difference between fed andstarved animals. Values are expressed asmeans � SE. n sizes are as follows. Experi-ment 1: gill (fed), 2; gill (starved) 3; mantleand retractor (fed) 5; mantle and retractor(starved) 7. Experiment 2: all conditions, 4,except cathepsin at pH 2.5, starved mantle,which is 3. Note: it is not possible to directlycompare measurements between experimentsdue to the use of different instrumentationand slightly different techniques.

Fig. 8. Maximal in vitro activity levels ofmetabolic enzymes in various tissues of fedand starved S. officinalis in experiment 1. Dataare expressed as �mol·min�1·mg protein�1.Solid symbols represent fed animals; opensymbols represent starved animals. Values areexpressed as means � SE. n sizes are as fol-lows: gill, 3 (in all cases); heart-aspartate ami-notransferase (AspAT) and alanine amino-transferase (AlAT), 3; heart - glutamate dehy-drogenase (GDH) and citrate synthase (CS), 5(fed) and 8 (starved); mantle and retractor, 4 inall cases, except CS, for which mantle is 2 andretractor is 3.

R434 PROTEIN DEGRADATION IN SEPIA

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00077.2012 • www.ajpregu.org

suggest activation of this process, which would result inproteolysis, leading to the production of amino acids to fuelmetabolism. Three studies with rainbow trout also report nochange in 20S proteasome activity in muscle following 14 daysof starvation, similar to the current finding with S. officinalis(9, 23, 35). Moreover, increases were reported in the expres-sion of ubiquitin-protein ligases (9, 35) and in polyubiquinatedprotein (35), consistent with activation of the UPS. AlthoughMartin et al. (23) present the argument, based primarily on adecrease in total ubiquitin protein, that 14 days of starvation inrainbow trout leads to a decrease in proteolysis by the UPS inmuscle, more recent findings do not support this viewpoint.The bulk of the evidence suggests that a period of starvation,resulting in almost total depletion of digestive gland triglycer-ide reserves in S. officinalis or a significant weight loss inrainbow trout, results in activation of the UPS in muscleinvolving increases in protein ubiquitination.

The situation in experiment 2, which is considered better toreflect the first responses to starvation, is quite different. In thiscase, there were no changes in any of the UPS variables inheart, mantle, and retractor muscle. In gill of starved asopposed to fed S. officinalis; however, 20S proteasome activityand polyubiquitin mRNA both significantly increased, as didthe average levels of proteasome �-subunit transcript andpolyubiquinated protein. It appears there is activation of theUPS in gill as an early response to starvation that could be partof gill remodeling discussed below.

Signaling molecules lead to tissue-specific responses. Thephosphorylation state of 4E-BP1 is considered to be an impor-tant controlling factor of protein synthesis. Dephosphorylated4E-BP1 binds to the eukaryote initiation factor, eIF4E andholds it in an inactive state. Upon phosphorylation, 4E-BP1pdissociates from eIF4E, allowing the latter to bind to eIF4Gand subsequently facilitate the onset of mRNA translation. Thephosphorylation state of 4E-BP1 is itself under control by Akt.Phosphorylation of Akt results in the phosphorylation of 4E-BP1 either directly or via a multistep process involving mTOR.Both pathways lead to the activation of protein synthesis (1,14). The decreased 4E-BP1p/4E-BP1 and Aktp/Akt that oc-curred in mantle in experiment 2 should work counter to theabove scenario and result in reduced rates of protein synthesisvia regulation of translation. We propose that in mantle, this is

an early response to a decrease in nutritional status and worksin parallel with the decrease in total RNA. The Akt response inS. officinalis mantle is similar to that previously described inmuscle of rainbow trout starved for 14 days (35), again illus-trating a common principle between a cephalopod and a te-leost. In starved rainbow trout, dephosphorylation of Akt isalso associated with increased expression of ubiquitin-proteinligases and polyubiquinated protein in muscle (35). In mam-malian tissues, downregulation of Akt signaling permits thetranscription of ubiquitin-protein ligases (34). A similar sig-naling mechanism is likely occurring in S. officinalis mantle aswell.

Once more the situation in gill, in which there was anincrease in 4E-BP1p/4E-BP1 ratio during the first stages offood deprivation, is strikingly different from the other tissues.The elevated phosphorylation of 4E-BP1 suggests an increasein mRNA translation and subsequent rates of protein synthesis.In concert with an activation of the UPS, this would result in atleast a transient increase in protein turnover.

Proteases and metabolic enzymes. The most commonlymeasured proteases cathepsin A-, B-, H- and L-like at pH5.5, cathepsin D- and E-like at pH 2.5, and calpain-like atneutral pH were all detectable in S. officinalis tissues. Thiswas not surprising since cathepsins have been well de-scribed in the viscera of S. officinalis (4). Activities of theseenzymes were generally higher in mantle and retractormuscles than gill. This is in contrast to 20S proteasomeactivity that is higher in gill than muscle tissues. Further-more, there is no obvious linkage between activity of theseproteases and protein synthesis as there is for 20S protea-some. The activity of these proteases were not influenced bystarvation with the one exception of cathepsin A-, B-, H-and L-like at pH 5.5 in mantle muscle. In this case, activityincreased in starved S. officinalis in experiment 1 whentriglyceride content in the digestive gland was lowest. Thisfinding suggests that the lysosomal pathway of proteindegradation is also activated to provide amino acids forenergy metabolism. Starvation in rainbow trout resulted inincreased expression of cathepsin L transcript in muscle(enzyme activity was not reported) (9) and activity ofcathepsin D in muscle of Atlantic cod (9, 16). Althoughthere may be minor variations on this theme, it appears thatthe lysosomal protein degradation pathway is activated atsome point during starvation in muscle of both S. officinalisand fishes.

Enzymes involved in amino acid trafficking (AspAT,AlAT, and GDH) and CS were detected in all tissues.Enzyme activities were always highest in heart reflecting thehigh energetic demands of the tissue. This finding is con-sistent with enzyme activity patterns in other cephalopods(2, 28). There were no differences in activity levels of anyof the metabolic enzymes between fed and starved S. offi-cinalis. These data though are only from experiment 1,which compared S. officinalis with intermediate to very lowlevels of triglyceride in the digestive gland. As such, al-though we think it is unlikely, we cannot rule out thepossibility that well-nourished specimens would exhibitdifferent activity levels of these enzymes.

As an aside to the main aspects of this study, we also reporta correlation between 20S proteasome and CS activities inheart. This relationship has also been noted for spotted wolffish

Fig. 9. 20S proteasome activity vs. CS activity in heart in experiment 1. Solidsymbols represent fed animals; open symbols represent starved animals. Theregression is significantly different from zero.

R435PROTEIN DEGRADATION IN SEPIA

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00077.2012 • www.ajpregu.org

(Anarhichas minor) and sea bass (Dicentrarchus labrax) (La-marre SG, unpublished data). The significance of this correla-tion is under investigation.

Perspectives and Significance

An initial response to food deprivation in mantle muscle ofS. officinalis, while triglyceride reserves are still available inthe digestive gland, appears to be a decrease in protein syn-thesis. This contention is based upon the dephosphorylation of4E-BP1, a signaling molecule intimately involved in the reg-ulation of translation, and phosphorylation of Akt that eitherdirectly or indirectly controls the phosphorylation state of4E-BP1. In addition, decreases in the level of total RNA wouldcontribute to a decrease in protein synthesis. As starvationprogresses to the point of almost total depletion of digestivegland triglyceride, protein degradation is activated. The proteinbreakdown through the UPS is enhanced via an increase inpolyubiquitinated protein available to the proteasome. This, inpart, would be due to an increase in ubiquitin, as the level ofpolyubiquitin mRNA is markedly elevated. In addition, in-creases in cathepsin activities imply that the lysosomal path-way for protein breakdown is also called upon. Together, theseprocesses would provide amino acids for energy metabolism,although increases in activity of specific transaminases are nota requirement for catabolism. This scenario is consistent withmodels presented for starvation in muscle of rainbow trout andother fishes.

The response to starvation in gill of S. officinalis is muchmore provocative and unlike that of mantle. In gill thereappears to be an increase in protein synthesis during the earlystages of food deprivation, as suggested by increased phos-phorylation of 4E-BP1. As well, levels of total RNA are higherin starved than in fed S. officinalis at similar levels of digestivegland triglyceride. At the same time, apparent rates of proteindegradation are enhanced through the UPS, as implied by asignificant increase in 20S proteasome activity and polyubiq-uitin mRNA, and a similar directional change in mean valuesof proteasome mRNA and polyubiquitinated protein. Takentogether, these changes imply enhanced protein turnover withmodification of the protein pool (i.e., gill remodeling). Changein gill structure occurs in fishes under various conditions. Forinstance, in carp (Carassius carassius) and goldfish (C. aura-tus) exposed to low temperature, with a subsequent decrease inoxygen consumption, there is increased mitotic activity and acell mass fills the space between lamellae to limit ion loss (37).In food-deprived juvenile S. officinalis, oxygen consumptionwas maintained for 6 days, but thereafter, it decreased to 35%of initial level by day 27. Following resumption of feeding,oxygen consumption immediately returned to control levels(15), revealing a maintenance of functional integrity of theoxygen delivery system. During the period of food deprivationassociated with reduced oxygen consumption, we propose thatthere is a reduction in requirement for exposed lamellae. This,in turn, could lead to gill remodeling, involving elevatedprotein turnover, perhaps to reduce the loss of a hemolymph-borne component normally obtained in the diet. Once feedingcommences, the oxygen consumption may be elevated, inpart, due to the selective protection of gill structure andfunction. Maintenance of gill function would also allow theanimal to meet the challenges of ionic and acid-base balance

with resumption of feeding and associated activity andcardiac output.

ACKNOWLEDGMENTS

We thank Connie Short and Kathy Clow for their technical assistance.

GRANTS

This work was supported in part by a Natural Sciences and EngineeringResearch Council (NSERC) of Canada Discovery Grant (to W. R. Driedzic),by the Centre National de la Recherche Scientifique and the UniversitéMontpellier 2. W. R. Driedzic holds the Canada Research Chair in MarineBioscience. S. G. Lamarre was supported by a postdoctoral fellowship from theCanadian Institutes for Health Research. D. Ditlecadet was supported by apostgraduate scholarship from NSERC.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: S.G.L., D.D., D.J.M., L.B., and W.R.D. conceptionand design of research; S.G.L., D.D., and W.R.D. performed experiments;S.G.L., D.D., and W.R.D. analyzed data; S.G.L., D.D., D.J.M., L.B., andW.R.D. interpreted results of experiments; S.G.L., D.D., D.J.M., L.B., andW.R.D. edited and revised manuscript; S.G.L., D.D., D.J.M., L.B., and W.R.D.approved final version of manuscript; W.R.D. prepared figures; W.R.D. draftedmanuscript.

REFERENCES

1. Balakuntalam SK, Mariappan MM, Sataranatarajan K, Lee MJ,Feliers D. mRNA translation: unexplored territory in renal science. J AmSoc Nephrol 17: 3281–3292. 2006.

2. Ballantyne JS, Hochachka PW, Mommsen TP. Studies on the metab-olism of the migratory squid, Loligo opalescens: enzymes of tissues andheart mitochondria. Mar Biol Lett 2: 75–85. 1981.

3. Bassaglia Y, Bekel T, Da Silva C, Poulain J, Andouche A, Navet S,Bonnaud L. ESTs library from embryonic stages reveals tubulin andreflectin diversity in Sepia officinalis (Mollusca - Cephalopoda). Gene498: 203–211. 2012.

4. Bihan EL, Zatylny C, Perrin A, Koueta N. Post-mortem changes inviscera of cuttlefish Sepia officinalis L. during storage at two differenttemperatures. Food Chem 98: 39–51, 2006.

5. Boucher-Rodoni R, Mangold K. Ammonia excretion during feeding andstarvation in Octopus vulgaris. Mar Biol 86, 193–197, 1985.

6. Carter CG, Lynch KA, Moltschaniwskyj NA. Protein synthesis in asolitary benthic cephalopod, the Southern dumpling squid (Euprymnatasmanica). Comp Biochem Physiol 153A: 185–190, 2009.

7. Castro BG, Garrido JL, Sotelo CG. Changes in composition of digestivegland and mantle muscle of the cuttlefish Sepia officinalis during starva-tion. Mar Biol 114: 11–20, 1992.

8. Clarke A, Rodhouse RG, Holmes LJ, Pascoe PL. Growth rates andnucleic acid ratio in cultured cuttlefish Sepia officinalis (Mollusca: Cepha-lopods). J Exp Mar Biol Ecol 133: 229–240, 1989.

9. Cleveland BM, Weber GM, Blemings KP, Silverstein JT. Insulin-likegrowth factor-I and genetic effects on indexes of protein degradation inresponse to feed deprivation in rainbow trout (Oncorhynchus mykiss). AmJ Physiol Regul Integr Comp Physiol 297: R1332–R1342, 2009.

10. Collison KS, Maqbool Z, Saleh SM, Inglis A, Makhoul NJ, Bakheet R,Al-Johi M, Al-Rabiah R, Zaidi M., Z, Futwan A, Al-Mohanna FA.Effect of dietary monosodium glutamate on trans fat-induced nonalcoholicfatty liver disease. J Lipid Res 50: 1521–1537, 2009.

11. Driedzic WR. Matching of cardiac oxygen delivery and fuel supply toenergy demand in teleosts and cephalopods. Can J Zool 66: 1078–1083,1988.

12. Driedzic WR, Sidell BD, Stewart JM, Johnston IA. Maximal activitiesof enzymes of energy metabolism in cephalopod systemic and branchialhearts. Physiol Zool 63: 615–629, 1990.

13. Foster AR, Houlihan DF, Hall SJ, Burren LJ. The effects of tempera-ture acclimation on protein synthesis and nucleic acid content on juvenilecod (Gadus morhua L). Can J Zool 70: 2095–2102. 1992.

R436 PROTEIN DEGRADATION IN SEPIA

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00077.2012 • www.ajpregu.org

14. Goodman CA, Mayhew DL, Hornberger TA. Recent progress towardunderstanding the molecular mechanisms that regulate skeletal musclemass. Cell Signal 23: 1896–1906, 2011.

15. Grigoriou P, Richardson CA. Effect of body mass, temperature and fooddeprivation on oxygen consumption rate of common cuttlefish, Sepiaofficinalis. Mar Biol 156: 2437–2481, (2009).

16. Guderly H, Lapointe D, Bédard M, Dutil JD. Metabolic prioritiesduring starvation: enzymes sparing in liver and white muscle of Atlanticcod, Gadus morhua L. Comp Biochem Physiol A 135: 347–356, 2003.

17. Hochachka PW, Fields JHA. Arginine, glutamate, and proline as sub-strates for oxidation and for glycogenesis in cephalopod tissues. PacificSci 36: 325–335, 1982.

18. Houlihan DF, Kelly K, Boyle PR. Correlates of growth and feeding inlaboratory-maintained Eledone cirrhosa (Cepalopoda: Octopoda). J MarBiol Assoc UK 78: 919–932, 1998.

19. Houlihan DF, McMillan DN, Agnisola C, Genoino IT, Foi L. Proteinsynthesis and growth in Octopus vulgaris. Mar Biol 106: 251–259, 1990.

20. Houlihan DF, McMillan DN, Laurent P. Growth rates, protein synthe-sis, and protein degradation rates in rainbow trout: effects of body size.Physiol Zool 59: 482–493, 1986.

21. Jones LJ, Upson RH, Haugland RP, Panchuk-Voloshina N, Zhou M,Haugland RP. Quenched BODIPY dye-labeled casein substrates for theassay of protease activity by direct fluorescence measurement. AnalBiochem 251: 144–152, 1997.

22. Lewis JM, Driedzic WR. Tissue-specific changes in protein synthesisassociated with seasonal metabolic depression and recovery in the northtemperate labrid, Tautogolabrus adspersus. Am J Physiol Regul IntegrComp Physiol 293: R474–R481, 2007.

23. Martin SAM, Blaney S, Bowman AS, Houlihan DF. Ubiquitin-protea-some-dependent proteolysis in rainbow trout (Oncorhynchus mykiss):effect of food deprivation. Pflügers Arch 445: 257–266, 2002.

24. Medina R, Wing SS, Goldberg AL. Increase in levels of polyubiquitinand proteasome mRNA in skeletal muscle during starvation and denerva-tion atrophy. Biochem J 307: 631–637, 1995.

25. Melzner F, Forsythe JW, Lee Wood JB PG, Piatkowski U, ClemmesenC. Estimating recent growth in the cuttlefish Sepia officinalis: are nucleicacid-based indicators for growth and condition the method of choice? JExp Mar Biol Ecol 317: 35–51, 2005.

26. Moltschaniwskyj NA. Understanding the process of growth in cephalo-pods. Mar Freshwater Res 55: 379–386, 2004.

27. Moltschaniwskyj NA, Carter CG. Protein synthesis, degradation, andretention: mechanisms of indeterminate growth in cepalopods. PhysiolBiochem Zool 83: 997–1008, 2010.

28. Mommsen TP, Hochachka PW, French CJ. Metabolism of arginine,proline, and ornithine in tissues of the squid, Illex illecebrosus. Can J Zool61: 1835–1846, 1983.

29. Mycek JM. Cathepsins. Methods Enzymol 19: 285–315, 1970.30. O’Dor R, Webber D. The constraints on cephalopods: why squid aren’t

fish. Can J Zool 64: 1591–1605, 1986.31. Packard A. Cephalopods and fish: limits of convergence. Biol Rev

Cambridge Philos Soc 47: 241–307, 1972.32. Pfaffl MW. A new mathematical model for relative quantification in

real-time RT-PCR. Nucleic Acids Res 29: 2003–2007, 2001.33. Rosa R, Pereira J, Nunes ML. Biochemical composition of cephalopods

with different life strategies, with special reference to a giant squid.Architeuthis sp. Mar Biol 146: 739–751, 2005.

34. Sandri M, Sandri C, Gilbert A, Skurk C, Calabria E, Picard A, WalshK, Schiaffino S, Lecker SH, Goldberg AL. Foxo transcription factorsinduce the atrophy-related ubiquitin ligase atrogin-1 and cause skeletalmuscle atrophy. Cell 117: 399–412, 2004.

35. Seiliez I, Panserat S, Skiba-Cassy S, Fricot A, Vachot C, Kaushik S,Tesseraud S. Feeding status regulates the polyubiquitination step of theubiquitin-proteasome-dependent proteolysis in rainbow trout (On-corhynchus mykiss) muscle. J Nutr 138: 487–491, 2008.

36. Shibatani T, Ward WF. Sodium dodecyl-sulfate (Sds) activation of the20s proteasome in rat-liver. Arch Biochem Biophys 321: 160–166, 1995.

37. Sollid J, Nilsson GE. Plasticity of respiratory structures. Adaptive remod-eling of fish gills induced by ambient oxygen and temperature. RespirPhysiol Neurobiol 154: 241–251, 2006.

38. Thompson VF, Saldaña S, Cong J, Goll DE. A BODIPY fluorescentmicroplate assay for measuring activity of calpains and other proteases.Anal Biochem 279: 170–178, 2000.

39. Vidal ÉAG, DiMarco P, Lee P. Effects of starvation and recovery on thesurvival, growth, and RNA/DNA ratio in loliginid squid paralarvae.Aquaculture 260: 94–105, 2006.

40. Wing SS, Haas AL, Goldberg AL. Increase in ubiquitin-protein conju-gates concomitant with the increase in proteolysis in rat skeletal muscleduring starvation and atrophy denervation. Biochem J 307: 639–645,1995.

R437PROTEIN DEGRADATION IN SEPIA

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00077.2012 • www.ajpregu.org