Proteomic responses of blue mussel (Mytilus) congeners to temperature acclimation

Temperature strongly affects rates of physiological processes and the integrity of macromolecular structures in cells, and thus plays an important role in limiting the range species can occupy (Hochachka and Somero, 2002; Pörtner, 2010; Tomanek, 2008; Tomanek, 2010). Increasing average global surface temperatures over the last three decades (0.6°C) (Hansen et al., 2006) correlate with a number of recently documented range shifts that are generally consistent with a temperature-driven range expansion of warm-adapted and range contraction of cold-adapted species at temperate northern latitudes, in both terrestrial (Parmesan et al., 1999; Thomas and Lennon, 1999) and marine (Barry et al., 1995; Mieszkowska et al., 2006) habitats. In order to predict which organisms will be affected most by anthropogenic temperature increases, we need to better understand the mechanisms by which temperature affects the physiology of organisms and, specifically, identify the cellular processes that are most sensitive to acute and chronic heat stress.

To address these questions, comparisons of closely related species (e.g. congeners) that differ in thermal tolerance and show recent range shifts may be especially useful. For example, the mussel genus Mytilus has been examined by a number of researchers, with several studies documenting recent range shifts by M. edulis along the Atlantic coast of North America (Jones et al., 2009), and by M. trossulus and M. galloprovincialis along the Pacific coast of North America (Sarver and Foltz, 1993; Wonham, 2004; Hilbish et al., 2010). Although these range shifts are likely to be affected by a number of biotic and abiotic factors, e.g. salinity (Braby and Somero, 2006a; Jones et al., 2009; Schneider, 2008), the above studies indicate that temperature is a crucial feature setting latitudinal range boundaries for each of the Mytilus species.

The M. trossulus–M. galloprovincialis species pair is especially informative because these congeners are closely related and the latter is an invasive species in the eastern Pacific, which is replacing the former throughout much of its original range in what appears to be a temperature-dependent fashion. Mytilus galloprovincialis, native to the Mediterranean, is thought to have first invaded southern California at the beginning of the last century (Geller, 1999). Since then it has displaced the native M. trossulus in the southern portion of its historical range, from Baja California north to central California, where a hybrid zone marks the current southern limit of M. trossulus. Single-species populations of M. trossulus are found north of the hybrid region into the Arctic, although isolated populations of M. galloprovincialis occur in Washington, perhaps as a result of independent invasion events. These differences in ranges, combined with studies examining the physiological (Schneider, 2008; Braby and Somero, 2006) and biochemical (Hofmann and Somero, 1996; Fields et al., 2006) responses of these congeners to increasing temperature, suggest that M. galloprovincialis is better able to compete in warmer conditions whereas M. trossulus is being restricted to regions with cooler temperatures.

Researchers have used congeners of the blue mussel complex to investigate changes in proteins in response to chronic temperature stress and have shown acclimation- and acclimatization-induced shifts in the synthesis of heat shock proteins (Hsps) (Hofmann and Somero, 1996) and pyruvate kinase, and the phosphorylation of stress-activated protein kinases (Anestis et al., 2007; Ioannou et al., 2009; Evans and Somero, 2010). However, to more fully describe the physiological differences between these Mytilus congeners that may provide M. galloprovincialis a competitive advantage over M. trossulus in warmer habitats, recent studies have taken a systems biology approach. For example, complementary studies of the transcriptomic (Lockwood et al., 2010) and proteomic (Tomanek and Zuzow, 2010) responses of both species to acute heat stress have provided a more comprehensive understanding of cellular responses to heat stress generally, as well as the differences in response between these two congeners.

Although proteomics studies generally are limited to the detection of medium- to high-abundance and mostly hydrophilic proteins (Lovric, 2010), recent comparative proteomics studies have generated a number of new hypotheses about how organisms respond to environmental stress, for example, how post-translational modifications may regulate the cellular response to oxidative stress (for reviews, see Sheehan, 2007; Tomanek, 2011). Proteomics studies on mollusks in particular have characterized the response to oxidative stress and pollutant exposure (Apraiz et al., 2006; McDonagh and Sheehan, 2007) as well as hypercapnia (Tomanek et al., 2011). Here we employ a proteomic analysis to better understand cellular responses to chronic temperature stress, because the acute diurnal fluctuations in temperature the Mytilus congeners experience are overlaid on a seasonal pattern of temperature variation. As anthropogenic global warming increases in rate and magnitude, intertidal organisms such as Mytilus will need to cope with stresses on multiple time scales, and changes in habitat temperature may alter the competitive balance between the native and the invasive species. The present study characterizes the proteomic response of M. trossulus and M. galloprovincialis to chronic temperature acclimation during constant immersion in the laboratory. We use two-dimensional (2-D) gel electrophoresis to separate proteins purified from gill tissue of temperature-acclimated M. trossulus and M. galloprovincialis individuals, and identify via mass spectrometry those proteins that change in abundance after acclimation.

Mytilus trossulus Gould 1850 and M. galloprovincialis Lamarck 1819 were collected subtidally from Newport, OR, USA (44°38′25″N, 124°03′10″W) and Santa Barbara, CA, USA (34°24′15″N, 119°41′30″W), respectively. In a separate study, PCR was used to confirm that each site was occupied by only a single species (i.e. there were no hybrids present) (Lockwood et al., 2010). Animals were kept for 4 weeks under constant immersion at 7, 13 or 19°C in recirculating seawater tanks and fed a phytoplankton diet every day. Following acclimation, gill tissue of both congeners was collected and immediately frozen using liquid nitrogen (N=6 per species for each treatment). Tissues were subsequently kept at –80°C until homogenization.

Sample preparation followed procedures outlined previously (Tomanek and Zuzow, 2010). Briefly, gill tissue was lysed in homogenization buffer [7 mol l^–1 urea, 2 mol l^–1 thiourea, 1% amidosulfobetaine (ASB)-14, 40 mmol l^–1 Tris-base, 0.5% immobilized pH 4–7 gradient (IPG) buffer (GE Healthcare, Piscataway, NJ, USA) and 40 mmol l^–1 dithiothreitol] at a ratio of 1:4. After centrifugation at 20°C for 30 min at 16,100 g, the proteins were precipitated by adding four volumes of ice-cold 10% trichloroacetic acid in acetone and incubating the solution at –20°C overnight. The precipitate was centrifuged at 4°C for 15 min at 18,000 g, the supernatant was discarded and the protein pellet was washed with ice-cold acetone and centrifuged again at 4°C. After air-drying, the pellet was re-suspended in rehydration buffer [7 mol l^–1 urea, 2 mol l^–1 thiourea, 2% cholamidopropyl-dimethylammonio-propanesulfonic acid (CHAPS), 2% nonyl phenoxylpolyethoxylethanol (NP)-40, 0.002% Bromophenol Blue, 0.5% IPG buffer and 100 mmol l^–1 dithioerythritol] through repeated vortexing. The protein concentration was determined with the 2-D Quant kit (GE Healthcare), according to the manufacturer’s instructions.

Proteins were separated by isoelectric point using IPG strips (pH 4–7, 11 cm; GE Healthcare). Samples were diluted to 2 μg μl^–1 with rehydration buffer, and 200 μl was loaded onto each strip in one well of an isoelectric focusing cell (BioRad, Hercules, CA, USA) via 5 h of passive rehydration, followed by 12 h of active rehydration (50 V). Separation and focusing was accomplished by exposing strips to 500 V for 1 h, 1000 V for 1 h and 8000 V for 2.5 h (all voltage changes occurred in rapid mode; maximum current 50 μA). Following focusing, strips were frozen at –80°C.

To prepare for 2nd dimension SDS-PAGE electrophoresis, strips were incubated in equilibration buffer (375 mmol l^–1 Tris-base, 6 mol l^–1 urea, 30% glycerol, 2% SDS and 0.002% Bromophenol Blue) for two 15 min intervals, first with 65 mmol l^–1 dithiothreitol and second with 135 mmol l^–1 iodoacetamide. IPG strips then were placed on top of 11.8% polyacrylamide gels, which were run (Criterion Dodeca; BioRad) at 200 V for 55 min at 10°C. Gels were subsequently stained with colloidal Coomassie Blue (G-250) overnight and destained by washing repeatedly with Milli-Q water for 48 h. The resulting gel images were scanned with an Epson 1280 transparency scanner (Epson, Long Beach, CA, USA).

Digitized images of 2-D gels were analyzed using Delta2D (version 3.6; Decodon, Greifswald, Germany) (Berth et al., 2007). Spots were detected on gels by fusing all images from one species into a composite image (Fig. 1). Spot boundaries were detected on this composite image and transferred back to the original gel images. After background subtraction, the relative amount of protein in each spot (i.e. spot volume) was quantified by normalizing against total spot volume of all proteins in the gel image.

To determine which proteins changed in volume significantly in response to acclimation temperature, we used a one-way ANOVA (P<0.02) within each species and with temperature as the main effect. We generated a null distribution for the one-way ANOVA (1000 permutations) to account for the unequal variance and non-normal distributions of the response variables. We chose a P-value of 0.02 to limit the number of false positives instead of using a multiple-comparison correction. Because there is only limited overlap between the proteome maps of the two congeners, as well as uncertainty whether overlapping proteins were orthologous or paralogous homologs, a two-way ANOVA comparing species was not possible. Following the one-way ANOVA, post hoc testing to compare treatments was conducted using Tukey’s analysis (P<0.05), using MiniTab (version 15; Minitab Inc., State College, PA, USA), to support conclusions about differences in single protein expression profiles (single-protein graphs are not shown).

Proteins that changed in abundance in response to temperature acclimation were excised from gels. Gel plugs were destained twice with 25 mmol l^–1 ammonium bicarbonate in 50% acetonitrile, dehydrated with 100% acetonitrile and digested with 11 ng μl^–1 trypsin (Promega, Madison, WI, USA) overnight at 37°C. Digested proteins were extracted using elution buffer [0.1% trifluoroacetic acid (TFA)/acetonitrile; 2:1] and concentrated using a SpeedVac centrifugal concentrator (Thermo Fisher Scientific, Waltham, MA, USA). The elution buffer containing the digested protein was mixed with 5 μl of matrix solution (0.2 mg ml^–1 α-hydroxycyano cinnamic acid in acetonitrile) and spotted on an Anchorchip™ target plate (Bruker Daltonics Inc., Billerica, MA, USA). The spotted proteins were washed with 0.1% TFA and recrystallized using an acetone/ethanol/0.1% TFA (6:3:1) mixture.

We obtained peptide mass fingerprints (PMFs) using a matrix-assisted laser desorption ionization tandem time-of-flight (MALDI-TOF-TOF) mass spectrometer (Ultraflex II; Bruker Daltonics Inc.). We selected a minimum of six peptides for tandem mass spectrometry in order to obtain information about the b- and y-ions of the peptide sequence for use in subsequent protein identification. Analysis of peptide spectra followed previously published procedures (Tomanek and Zuzow, 2010). We used flexAnalysis (version 3.0; Bruker Daltonics Inc.) to detect peptide peaks (SNAP algorithm with a signal-to-noise ratio of 6 for MS and 1.5 for MS/MS). Porcine trypsin was used for internal mass calibration.

To identify proteins we used Mascot (version 2.2; Matrix Science Inc., Boston, MA, USA) and combined PMFs and tandem mass spectra in a search against two databases. One database is an EST library that initially contained approximately 26,000 entries, representing 12,961 and 1688 different gene sequences for M. californianus and M. galloprovincialis, respectively (Lockwood et al., 2010). The other was Swiss-Prot (June 2009), with 17,360 molluscan protein sequences. Oxidation of methionine and carbamidomethylation of cysteine were our only variable modifications. Our search allowed one missed cleavage during trypsin digestion. For tandem mass spectrometry we set the precursor-ion mass tolerance to 0.6 Da, the default value in Mascot. The molecular weight search (MOWSE) score that indicated a significant hit was dependent on the database: scores higher than 40 and 51 were significant (P<0.05) for a search in the Mytilus EST and Swiss-Prot databases, respectively. However, we only accepted positive identifications that included two matched peptides regardless of the MOWSE score.

To associate proteins with similar expression patterns across samples, and to group mussels with similar protein expression profiles, we employed hierarchical clustering with average linking (Delta2D) using a Pearson correlation metric. To further assess the importance of specific proteins in differentiating the proteomes of mussels acclimated to different temperatures, we also employed principal component analyses (PCAs; Delta2D); these PCAs were based on proteins whose expression profiles changed significantly after acclimation and that were identified with mass spectrometry (ANOVA, P<0.02). (PCAs were also performed for each species using all spots detected, and the separation of acclimation groups in each of these PCAs was not qualitatively different from that in PCAs that included only identified spots.) Component loadings, which quantify the contribution of each protein to the separation of samples along a given component, were used to compare changes in protein expression both among temperatures and between species.

The results described below are based on changes in protein abundance in gill of M. trossulus and M. galloprovincialis after acclimation to different temperatures. It is important to note that protein abundance can change because of synthesis, post-translational modifications (PTMs) or degradation. Thus, when we refer to abundance changes of protein spots or changes in protein expression, one or a combination of these processes may be responsible.

Following acclimation of mussels to constant immersion at 7, 13 and 19°C, we separated protein samples from gill tissue of both species via 2-D gel electrophoresis (2DGE) (Fig. 1). We detected 331 and 458 protein spots from M. trossulus and M. galloprovincialis, respectively. Of these, 23.9% (79 spots) changed in abundance among M. trossulus acclimated to different temperatures, of which 33 spots were identified via MS/MS (supplementary material Table S1). In comparison, 32.8% (150 spots) changed in M. galloprovincialis in response to temperature acclimation, of which we identified 48 (supplementary material Table S2).

Fig. 2 shows the hierarchical clustering of M. trossulus samples (i.e. individual mussels, arrayed in columns) and of identified protein isoforms (arrayed in rows) that significantly (ANOVA, P<0.02) differ in abundance following temperature acclimation. Fig. 3 shows the comparable hierarchical clustering of samples and proteins from M. galloprovincialis. For each species, samples cluster clearly by acclimation temperature, indicating that acclimation to each of the three temperatures significantly altered protein expression profiles.

Examination of Figs 2 and 3 reveals that whereas M. trossulus acclimated to 7 and 13°C have roughly comparable protein expression patterns, i.e. there are only two major protein clusters that differentiate 7 and 13°C together from 19°C (clusters 1–2; Fig. 2), M. galloprovincialis proteins cluster in four major groups (clusters 1–4; Fig. 3), with the 7, 13 and 19°C acclimated groups having distinct expression profiles.

Clusters 1 and 2 are comparable between the two species: cluster 1 includes proteins upregulated at 7 and 13°C but downregulated at 19°C; cluster 2 includes proteins with the reverse pattern. The latter cluster is characterized in both species by the upregulation at 19°C of cytoskeletal proteins, e.g. actin and tubulin, as well as energy-generating proteins of the electron transport chain (ETC), e.g. ubiquinol cytochrome c reductase. In the case of M. trossulus, the relatively cold-adapted congener, this cluster also includes the upregulation at 19°C of molecular chaperones, e.g. Hsp25, Hsp70 and glucose-regulated protein 78 (Grp78), and one oxidative stress protein, Dyp-type peroxidase. However, in M. trossulus, each of these proteins has two or more isoforms, one that is upregulated at 19°C and at least one other that is upregulated at 7 and 13°C (Fig. 2). Inversely correlated changes in abundance of isoforms of the same protein may be due to differing PTMs or combinations of PTMs (Walsh et al., 2005) that alter the positions of the protein on the 2-D gel, and this is our interpretation of the changing abundance of Hsps in M. trossulus after temperature acclimation. However, we cannot determine from abundance data whether a particular shift in position on the 2-D gel, due to a specific PTM, leads to an increase or decrease in chaperoning activity. One Hsp25 isoform in M. trossulus (spot 18) is notable because its molecular mass estimated from the 2-D gel is more than double the predicted mass (supplementary material Table S1).

Cluster 1, characterized by the upregulation of proteins at 7 and 13°C, shows fewer cytoskeletal proteins but a number of chaperones (five in M. trossulus, and three in M. galloprovincialis, including cyclophilin 3, T complex protein-1 and prohibitin), proteins involved in energy metabolism, e.g. pyruvate dehydrogenase (in M. galloprovincialis) and isocitrate dehydrogenase (M. trossulus), and several oxidative stress proteins (see below for discussion), e.g. Dyp-type peroxidase (M. trossulus), and thioredoxin peroxidase and carbonyl reductase (M. galloprovincialis).

The two species are differentiated, however, by the presence of two additional protein clusters that are found only in M. galloprovincialis (clusters 3 and 4; Fig. 3) and that separate mussels acclimated to 7°C from those acclimated to 13°C. The presence of these clusters suggests that M. galloprovincialis, the relatively warm-adapted congener, is more sensitive to the difference between these lower temperatures than M. trossulus. Cluster 3, which includes proteins with relatively high abundance at 7 and 19°C but low abundance at 13°C, is the smallest cluster, with only five proteins. These five proteins are members of varied functional groups, including some (major vault protein, fibrinogen-like domain) whose function is unclear. Cluster 4 comprises proteins upregulated at 13°C relative to the other two treatment temperatures, and is significantly larger than cluster 3, with 14 proteins (29% of identified proteins). This cluster is dominated by cytoskeletal proteins (10 of 14), but includes two proteins associated with energy generation as well as a small Hsp.

PCAs were conducted to further assess the impact of the three acclimation temperatures on protein expression in the Mytilus congeners. Generally, the outcome of the PCAs supports the results described above for our hierarchical cluster analysis – both species show a significant change in protein expression pattern in response to acclimation to 19°C, but M. galloprovincialis is more strongly affected by the difference in temperature between the 7°C and 13°C acclimation treatments. Note that because of the variation between the proteomes of these congeners, it was not possible to overlap protein spots and assume homologies, thus we conducted a separate PCA for each species (Fig. 4).

In both species, the first component (horizontal axis) differentiates the 7°C and 13°C acclimation groups from the 19°C group; in M. trossulus (Fig. 4A), PC1 explains 61.5% of the variation in expression of the 33 identified proteins, whereas in M. galloprovincialis (Fig. 4B), PC1 explains 55.9% of the variation among the 48 identified protein spots. The second component (vertical axis) most clearly separates the 7°C and 13°C acclimation groups in both species. Notably, however, PC2 in M. trossulus accounts for only 12.9% of protein expression variation, but for M. galloprovincialis PC2 encompasses approximately double that amount, 25.4%.

The component loadings associated with these PCAs quantify the contribution of each protein to the variation explained by each PC. For example, in the case of PC1, proteins with highly positive loadings are relatively upregulated in the 19°C groups and proteins with highly negative loadings are relatively upregulated in the 7°C and 13°C groups. Among the proteins with the 10 highest positive loadings for PC1 in M. galloprovincialis, only proteins that are part of the cytoskeleton or are involved in energy metabolism are found (Table 1A). In contrast, in M. trossulus, the highest positive loadings for PC1 comprise four proteins of the cytoskeleton, including gelsolin, two molecular chaperones and three proteins involved in energy metabolism as well as one oxidative stress protein, Dyp-type peroxidase (Table 1A). Negative loadings for M. galloprovincialis on PC1, which indicate proteins that were upregulated at 7 and 13°C relative to 19°C, include five cytoskeletal proteins, one protein involved in energy metabolism, two molecular chaperones and one oxidative stress protein (Table 1B). The possible role of prolyl 4-hydroxylase will be discussed below. A similar pattern is seen for M. trossulus (Table 1B), although proteins highly upregulated after cold acclimation in this species include three putative oxidative stress proteins.

The second component most significantly separates the 7°C from the 13°C treatment group in both species (Fig. 4). Along PC2, M. trossulus has relatively few proteins with high loadings (either positive or negative) relative to M. galloprovincialis (Table 1C,D), indicative of the small variance explained by this component in the former species, and supporting the relative insensitivity of M. trossulus to the difference between the 7°C and 13°C acclimation treatments demonstrated by the hierarchical cluster analysis (Fig. 2).

The highest positive loadings of the second component in the more cold-sensitive M. galloprovincialis, which indicate upregulation at 7°C relative to 13°C, include proteins with a range of cellular functions, including cellular stress (major vault protein), oxidative stress (NADP-dependent IDH and thioredoxin peroxidase) and protein homeostasis (a proteasome subunit and cyclophilin) (Table 1C). In M. trossulus, two isoforms of NADP-dependent IDH and Grp78 as well as tubulin have the highest positive loadings, with other proteins contributing relatively little to the variance in PC2 (Table 1C). The 10 proteins with the highest negative loadings of PC2 for M. galloprovincialis are mainly isoforms of tubulin and actin, with ATP synthase and esterase D being the exceptions (Table 1D). Again, M. trossulus has few proteins with high negative component loadings for PC2, with a value above 1.0 for only three proteins: a heavy metal binding protein, Hsp25 and a β-tubulin isoform (Table 1D). To reiterate, the component loadings of PC2 indicate that M. galloprovincialis responds strongly to the difference between 7 and 13°C, whereas M. trossulus does not.

A number of studies have provided evidence that M. galloprovincialis is the more warm-adapted (but cold-sensitive) of the two congeners (e.g. Braby and Somero, 2006; Schneider and Helmuth, 2007). Based on seasonal seawater temperature data (Ioannou et al., 2009), 7°C is a temperature M. galloprovincialis may rarely experience within its native (Mediterranean) environment. It can endure long-term exposure to 28°C, however, with 70% of individuals surviving this temperature for at least 30 days under laboratory conditions (Anestis et al., 2007). In contrast, 21°C is an acclimation temperature at which M. trossulus experiences high levels of mortality (Braby and Somero, 2006). Further, winter-acclimatized M. trossulus induced Hsp synthesis after acute exposure to 23°C (Buckley et al., 2001). At lower temperatures, M. trossulus acclimated to 7°C had a heart rate two to three times higher than that of M. galloprovincialis acclimated to the same temperature, and M. trossulus was able to maintain a higher heart rate when acutely exposed to lower temperatures (Braby and Somero, 2006). We can summarize, then, that 7°C appears to induce physiological stress in M. galloprovincialis, whereas 19°C may approach the long-term upper thermal limit of M. trossulus.

We know of few comparative proteomics studies examining effects of environmental stress on closely related species (Tomanek and Zuzow, 2010; Serafini et al., 2011); such studies have been uncommon in part because comparing proteomes between species poses several conceptual and technical challenges. First, even the proteomes of closely related species differ enough to prevent a simple matching of proteome maps, which limits our ability to apply a single PCA to the combined data set of both species. Second, protein abundance comparisons ideally should be between orthologous homologs with the same PTMs, which requires that the mass spectra cover a high percentage of each protein sequence to confirm close homology, as well as identify all the peptides that carry PTMs. Third, because a single amino acid change or PTM can alter the isoelectric point of a protein significantly, homologous protein isoforms between congeners may occur outside the pH range being analyzed (pH 4–7 in this study). Finally, because of the lack of a full genome sequence in Mytilus, it is possible that certain homologs are not included in the protein or nucleotide databases and thus cannot be identified. As a consequence, the comparison of proteomic data sets between species will not identify every pair of homologous proteins whose abundances vary; instead, our goal is to understand differences in metabolism and cytoskeletal composition between species, based on changes in abundance of as many identified proteins as possible.

Despite the apparent differences in physiological responses to higher temperatures described above, M. trossulus and M. galloprovincialis do share similarities in their proteomic response to chronic heat exposure. For example, in each species, a comparable cluster of proteins is upregulated at 19°C (cluster 2; Figs 2, 3). Tubulin isoforms, which are important components of this cluster, form microtubules that are the basic structural elements of cilia that in Mytilus drive water flow through the gills. This active movement of water is necessary for respiration as well as for transporting mucus containing trapped food particles over the epithelium and along the ventral groove towards the mouth. Both acute and chronic exposure to warmer temperatures in Mytilus increase the filtration rate and particle transport (i.e. velocity along the ventral grove) without signs of temperature compensation (Kittner and Riisgård, 2005; Rao, 1953; Richoux and Thompson, 2001), and thus may require synthesis or remodeling of cilia. The increase in filtration rate at higher temperatures also likely requires higher rates of ATP production, which may be achieved by increasing the abundance of enzymes of the ETC and oxidative phosphorylation (Aiello, 1960; Malanga and Aiello, 1972). These include ubiquinol cytochrome c reductase (complex III of the ETC), succinate dehydrogenase (complex II, as well as a component of the Krebs cycle) and ATP synthase β-subunit, which are upregulated at 19°C in one or both species (Figs 2, 3, Table 1A). Note, however, that NADH dehydrogenase (complex I of the ETC) and pyruvate dehydrogenase (linking glycolysis to the Krebs cycle) are not part of this cluster in M. galloprovincialis but instead are upregulated at 13°C and downregulated at 19°C (Fig. 3; see below for further discussion).

Notably, cluster 2 also includes two classes of proteins unique to M. trossulus – molecular chaperones and oxidative stress proteins. In M. trossulus only, acclimation to 19°C increased levels or shifted the position, possibly due to PTMs, of several molecular chaperones: Hsp70, Hsp25 and an isoform of glucose-regulated protein 78 [Grp78, also known as binding immunoglobulin protein (BiP), is the Hsp70 homolog of the endoplasmic reticulum] (Fig. 2). The high component loadings for two of these proteins along PC1 (Table 1A,B) indicate that changes in their expression are an important part of the acclimation response to 19°C in M. trossulus. In addition to heat stress, the increased abundance of one DyP-type peroxidase isoform (another one is upregulated at 7 and 13°C), an enzyme that functions similarly to catalase and catalyzes the degradation of H₂O₂ to H₂O (Sugano, 2009), may suggest higher levels of oxidative stress at 19°C in M. trossulus but not in M. galloprovincialis.

Several studies have quantified changes in Hsp abundance and synthesis in response to high temperature in Mytilus. Anestis and colleagues (Anestis et al., 2007) showed that Hsp70 levels in M. galloprovincialis increase following acclimation to 26°C, whereas during seasonal acclimatization increases in abundance seem to occur at lower temperatures, perhaps just above 20°C (Hamer et al., 2004; Ioannou et al., 2009; Minier et al., 2000). Buckley et al. (Buckley et al., 2001) examined acclimation and acclimatization in M. trossulus, and found a difference in Hsp70 induction temperature depending on season (23°C in mussels collected in winter, 28°C in summer). In the single proteomics study comparing acute heat stress in both congeners, M. trossulus induced one Hsp70 isoform at 24°C, and M. galloprovincialis induced a heat shock cognate 70 isoform at 24°C, but interspecific differences in patterns of chaperone expression were complex at higher temperatures (Tomanek and Zuzow, 2010). In summary, it does not seem surprising that M. galloprovincialis shows no increase in Hsp abundance during acclimation to 19°C. The fact that M. trossulus changed the abundances (or positions) of several Hsps at the same temperature suggests that it regulates its chaperone activity in response to acclimation at this temperature. If position shifts within the 2-D gels indeed are underlying the ‘up-and-down’ pattern of several chaperones during acclimation (Fig. 2), it will be important to determine whether and which PTMs are causing these shifts and how they might change protein activity.

Protein cluster 1 (Figs 2, 3) is characterized by the upregulation of proteins at 7 and 13°C relative to the acclimation to 19°C. In both species, the proteins of this cluster fall into three functional categories: cytoskeletal proteins, molecular chaperones and oxidative stress proteins. Compared with the warm-acclimation cluster 2, cold acclimation leads to an increase in protein isoforms involved in chaperone activity and oxidative stress and a decrease in many cytoskeletal protein isoforms.

Among those cytoskeletal proteins that changed expression levels, we found mainly actin and tubulin isoforms, as we did in cluster 2, but fewer of them. Some of the changing levels of tubulin and actin isoforms may be related to changes of T-complex protein-1 (also known as chaperonin containing TCP-1 or CCT) and radial spoke head 9 (Figs 2, 3). The mature conformation of both actin and tubulin requires the folding cage of TCP-1 (Sternlicht et al., 1993), a key chaperone that also plays a role in the regulation of the formation of mitotic spindles and thus cell proliferation (Brackley and Grantham, 2009). In M. galloprovincialis only, T-complex protein-1 was upregulated at 7 and 13°C relative to 19°C (Fig. 3, Table 1B), indicating that higher rates of folding of cytsokeletal proteins, possibly for the repair or generation of cilia, were required at these two temperatures in this warm-adapted species. Similarly, radial spoke head proteins are important components of the axoneme that forms the core of cilia. Radial spoke head 9 is upregulated in cluster 1 of both M. galloprovincialis and M. trossulus, suggesting increased synthesis or turnover of cilia at lower temperatures.

A second functional category present in cluster 1 includes molecular chaperones that are folding catalysts but are also involved in stabilizing proteins during unfolding due to proteotoxic stress (Hartl and Hayer-Hartl, 2002). In cold-acclimated M. trossulus, two isoforms of the major cellular chaperones Hsp70 and small Hsp (Hsp25) are upregulated (Fig. 2, Table 1B), with the latter playing an important role in stabilizing cytoskeletal elements during stress (Haslbeck et al., 2005). In addition, one isoform of Grp78 is upregulated (another, spot 79, is downregulated); the endoplasmic Grp78 is implicated, among other roles, in moderating the effects of oxidative stress (Liu et al., 1997). In M. galloprovincialis, the small Hsp21 shows upregulation in response to cold acclimation (Fig. 3), as does the β-subunit of prolyl 4-hydroxylase. Prolyl 4-hydroxylase holozyme modifies prolyl residues of collagens and several transcription factors, including hypoxia-inducible factor 1 (Gorres and Raines, 2010). Interestingly, though, the β-subunit of this heterotetramer is a protein disulfide isomerase (Koivunen et al., 2005; Vranka et al., 2004), whose role is to assist in the folding of proteins through the formation of disulfide bonds in the endoplasmic reticulum (Malhotra and Kaufman, 2007).

The third functional category in cluster 1 includes proteins involved in scavenging reactive oxygen species (ROS). For example, isoforms of NADP-dependent isocitrate dehydrogenase are upregulated during cold acclimation in both species. NADP-dependent IDH, an enzyme linked to the Krebs cycle, increases the [NADPH]/[NADP] ratio in the mitochondrion; the repair of macromolecules that suffer oxidative damage is dependent on high [NADPH], which is used to re-reduce glutathione (Jo et al., 2001). NADP-dependent IDH also has been hypothesized to play a role in the antioxidant response of Mytilus during acute heat stress (Tomanek and Zuzow, 2010).

Some oxidative stress proteins upregulated after cold acclimation are unique to each of the congeners. Examples in M. galloprovincialis include thioredoxin peroxidase (also known as peroxiredoxin), which detoxifies hydrogen peroxide or organic peroxides (Cox et al., 2010), and carbonyl reductase, which metabolizes lipid peroxidation products like carbonyl-containing aldehydes (Ellis, 2007). The function of selenium binding protein-1 is not known, but it may interact with glutathione peroxidase and thus affect the scavenging of ROS indirectly (Fang et al., 2010). In M. trossulus, DyP-type peroxidase catalyzes the conversion of H₂O₂ into H₂O (Sugano, 2009).

By reacting with unsaturated fatty acids, ROS produce many types of aldehydes, among the most common being malondialdehyde, hexanal and 4-hydroxynonenal (Ellis, 2007). Notably, according to the concept of homeoviscous adaptation, the levels of unsaturated phospholipids in membranes increase during cold acclimation to maintain membrane fluidity (Hazel and Williams, 1990). Thus it is likely that the level of cytotoxic aldehydes will increase as temperature decreases. In turn, these aldehydes cause protein carbonylation, which affects protein conformation and function, leading to protein denaturation (McDonagh et al., 2005). Based on these changes, an important oxidative stress protein upregulated during cold acclimation in both species is an isoform of aldehyde dehydrogenase (ALDH2) located in the mitochondrion (Ellis, 2007). In M. galloprovincialis, ALDH2 is upregulated at 13°C (Fig. 3), but its positive loading on PC2 (Table 1C) indicates that its abundance increases even further from 13 to 7°C. In contrast, in M. trossulus, ALDH2 has a highly negative loading for PC1 but not PC2 (Table 1B), indicating that it is upregulated at 13°C relative to 19°C, but does not increase significantly in abundance as temperature decreases further to 7°C.

Our results agree with those of other studies examining evidence for an increasing rate of oxidative stress, including lipid peroxidation, during cold acclimation in ectotherms. For example, proteomic changes similar to those described here were shown in a study on the response to cold stress of livers of warm-adapted gilthead sea bream (Ibarz et al., 2010); a study on cultivation temperatures of wild-type C. elegans larvae also showed increasing levels of chaperones and IDH in response to cold (Madi et al., 2003).

In addition to the differences between the species in clusters 1 and 2, the presence of clusters 3 and 4 in M. galloprovincialis (Fig. 3) provides further evidence that the proteomic responses to acclimation to 7 and 13°C differ between the congeners. Tubulin isoforms are the most abundant protein in the clusters unique to M. galloprovincialis. They, together with ATP-synthase β-subunit, contributed the most to the negative loadings of PC2, i.e. those indicating upregulation at 13°C relative to 7°C (Table 1D; see also Fig. 3). One potential explanation for this proliferation of tubulin isoforms could lay in the upregulation of formylglutathione hydrolase at 13°C, an enzyme that produces glutathione. Protein glutathionylation is an important and widespread reaction to protect sulfhydryl groups (Dalle-Donne et al., 2009), specifically in tubulins (Landino et al., 2004).

Comparing the results of the present study with those of our previous analysis of proteomic changes during acute heat stress in the two Mytilus congeners (Tomanek and Zuzow, 2010) may offer insights into biochemical processes that are differently regulated between chronic and acute heat stress. For example, given that changes in tubulin can occur in clusters in which few or no concomitant changes in molecular chaperones are observed (13 and 19°C acclimation in M. galloprovincialis), changes in tubulin expression may not indicate thermal damage to the cytoskeletal network. Instead, changes in tubulin expression may correspond to changes in the activity of cilia due to increasing ventilation and filtration rates with increasing temperature (Kittner and Riisgård, 2005; Rao, 1953; Richoux and Thompson, 2001), or changing requirements for microtubules during cell proliferation. In contrast, during recovery from acute stress, changes in tubulin isoform expression occur in parallel with changes in chaperone abundance, especially in M. galloprovincialis (Tomanek and Zuzow, 2010).

In M. trossulus, we observed changes in a chaperone from the endoplasmic reticulum, Grp78, which did not change during acute heat stress (Tomanek and Zuzow, 2010). The total abundance of the two Grp78 isoforms (spots 76 and 79; data not shown) was greatest at 7 and 19°C, indicating that the endoplasmic reticulum of gill cells responds to both low and high temperature acclimation, but not to acute heat stress, with higher chaperoning capacity.

One pattern that appears to be common to both acute and chronic heat stress may represent a trade-off between the expression of oxidative stress proteins and proteins involved in energy metabolism. For example, during acclimation, both species show an upregulation of ubiquinol cytochrome c reductase (complex III of the ETC) at 19°C only. Complex III is a major site of ROS production (Andreyev et al., 2005), but oxidative stress proteins increase in abundance mostly at 7 and 13°C (Figs 2, 3; DyP-type peroxidase in M. trossulus being the only exception). At least in M. galloprovincialis, protein expression patterns may also indicate additional adjustments in energy metabolism, e.g. upregulation of succinate dehydrogenase and downregulation of NADH dehydrogenase at 19°C, downregulation of pyruvate dehydrogenase at 19°C, and upregulation of ATP-synthase β subunit isoforms at 13°C (spot 59) and 19°C (spot 143). These changes did not occur in M. galloprovincialis during acute heat stress and could indicate that chronic adjustments, due to cold-induced oxidative stress versus a heat-induced increase in ventilation rate, force different trade-offs between limiting oxidative stress and maintaining an appropriate level of ATP production.

Similarly, the NADP-dependent IDH isoforms we detected are part of the Krebs cycle and play an important regulatory role in maintaining a constant ratio of reduced to oxidized glutathione, which in turn controls the antioxidant potential of the cell (Jo et al., 2001). IDH isoforms changed abundances during acute heat stress in M. trossulus, consistent with an increasing role in scavenging ROS (Tomanek and Zuzow, 2010). Here, all IDH isoforms (spot 75 in M. galloprovincialis; spots 68 and 73 in M. trossulus) have high positive loadings for PC2 and are more abundant at 7°C relative to 13°C (although two of these isoforms, spots 75 in M. galloprovincialis and 68 in M. trossulus, are also more abundant at 19°C compared with 13°C, their PC1 loadings are negligible). In contrast to mitochondrial IDH expression, several enzymes of the cytosolic pentose phosphate pathway that provide reducing equivalents in form of NADPH during acute heat stress did not change during chronic temperature stress in either congener.

Overall, a broad comparison of the proteomic data sets derived from chronic and acute temperature stress suggests that oxidative stress levels seem to be high in chronically cold-acclimated as well as acutely heat-stressed mussels. Thus, the combination of cold acclimation (or acclimatization) followed by acute heat exposure could represent a particularly severe level of stress and be a major limiting factor in setting distribution ranges, a notion that is now supported by a number of physiological data sets (Tomanek, 2008; Tomanek, 2010).

In summary, both congeners experience cold-induced oxidative stress, with the relatively warm-adapted M. galloprovincialis showing different expression patterns between 7 and 13°C (possibly because of tubulin isoform expression) whereas the cold-adapted M. trossulus responds with a single expression pattern to both temperatures. Cold-induced oxidative stress also may cause re-arrangements in energy metabolism in both species, e.g. downregulation of ubiquinol cytochrome c reductase and ATP synthase, again with different expression patterns between 7 and 13°C in M. galloprovincialis only. At the warmest acclimation temperature, both species show an upregulation of several tubulin isoforms and ubiquinol cytochrome c reductase, in agreement with previous studies that showed that filtration and respiration rates increase with increasing temperature due to enhanced ciliary activity (Aiello, 1960; Gray, 1924). Furthermore, during acclimation to 19°C, M. trossulus, but not M. galloprovincialis, shows increasing abundances of molecular chaperones, suggesting that this temperature is close to its chronic thermal limit.

 
• 2DGE

  two-dimensional gel electrophoresis

 
• ADP (ATP)

  adenosine di-(tri)-phosphate

 
• ALDH

  aldehyde dehydrogenase

 
• EST

  expressed sequence tag

 
• ETC

  electron transport chain

 
• Grp78

  glucose-regulated protein 78

 
• Hsp

  heat shock protein

 
• IDH

  isocitrate dehydrogenase

 
• IPG

  immobilized pH gradient

 
• MALDI-TOF-TOF

  matrix-assisted laser desorption ionization tandem time-of-flight

 
• MOWSE

  molecular weight search

 
• MS/MS

  tandem mass spectrometry

 
• NAD(H)

  nicotinamide adenine dinucleotide (reduced form)

 
• NADP(H)

  nicotinamide adenine dinucleotide phosphate (reduced form)

 
• PC

  principal component

 
• PCA

  principal component analysis

 
• PMF

  peptide mass fingerprint

 
• PTM

  post-translational modification

 
• ROS

  reactive oxygen species

 
• TCP-1

  T-complex protein 1

 
• TFA

  trifluoroacetic acid