Functional Determinants of Temperature Adaptation in
Enzymes of Cold- versus Warm-Adapted Mussels (Genus
Mytilus)
Brent L. Lockwood*, ,1 and George N. Somero1
1
Hopkins Marine Station, Stanford University
Present address: Department of Biology, Indiana University
*Corresponding author: E-mail: [email protected].
Associate editor: John H. McDonald
Abstract
Key words: adaptation, invasive species, Michaelis–Menten constants (Km), Mytilus, protein evolution, structure–function
relationships, temperature.
Introduction
Temperature affects stabilities of protein 3D structures
(Somero 1995; Fields 2001), causing protein functions like
ligand recognition and binding to be thermally sensitive.
Therefore, natural selection should favor amino acid sequences that optimize protein function in a particular thermal
environment, for example, by conserving ligand-binding
abilities within narrow ranges (Jaenicke 1991; Fields 2001;
Berezovsky and Shakhnovich 2005). Accordingly, temperature adaptation at the protein level has been documented
across a diversity of ectothermic taxa inhabiting a wide range
of thermal habitats (Watt 1977; Place and Powers 1979; Fields
and Somero 1998; Fields et al. 2002; Watt et al. 2003; Bae and
Phillips 2004; Johns and Somero 2004; Somero 2004; Fields
et al. 2006; Siddiqui and Cavicchioli 2006; Dong and Somero
2009). Although many species have been studied, the variety
of different classes of enzymes examined in such studies remains limited. Many previous studies have found evidence
for temperature adaptation of functional properties, such as
the Michaelis–Menten constant (Km), in dehydrogenase
enzymes—proteins chosen for study because they are widely
occurring among species, abundant in the cell and relatively
straightforward to purify and assay (Place and Powers 1979;
Graves and Somero 1982; Fields and Somero 1998; Fields et al.
2006; Dong and Somero 2009). These enzymes have shared
features of their 3D structures, namely the NAD(P)-binding
Rossmann fold and large changes in conformation during ligand binding. However, a central question remains unanswered: to what extent does the thermal environment
affect the evolution of other structural classes of enzymes?
The blue mussel species complex is a powerful model
system for elucidating mechanisms of temperature adaptation (Braby and Somero 2006; Lockwood et al. 2010;
Tomanek and Zuzow 2010; Lockwood and Somero
2011) due to its well-characterized thermal evolutionary
history (Thunell 1979; Mann and Hamilton 1995) and recent divergence (ca. 3.5 Ma) of the three congeners, Mytilus
trossulus, Mytilus edulis, and Mytilus galloprovincialis
(Vermeij 1991; Seed 1992). In particular, M. galloprovincialis, a native of the Mediterranean Sea, has a warm-adapted
© The Author 2012. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please
e-mail: [email protected]
Mol. Biol. Evol. 29(10):3061–3070. 2012 doi:10.1093/molbev/mss111
Advance Access publication April 6, 2012
3061
Research article
Temperature is a strong selective force on the evolution of proteins due to its effects on higher orders of protein structure
and, thereby, on critical protein functions like ligand binding and catalysis. Comparisons among orthologous proteins from
differently thermally adapted species show consistent patterns of adaptive variation in function, but few studies have
examined functional adaptation among multiple structural families of proteins. Thus, with our present state of knowledge,
it is difficult to predict what fraction of the proteome will exhibit adaptive variation in the face of temperature increases of
a few to several degrees Celsius, that is, temperature increases of the magnitude predicted by models of global warming.
Here, we compared orthologous enzymes of the warm-adapted Mediterranean mussel Mytilus galloprovincialis and the
cold-adapted Mytilus trossulus, a native of the North Pacific Ocean, species whose physiologies exhibit significantly
different responses to temperature. We measured the effects of temperature on the kinetics (Michaelis–Menten
constant—Km) of five enzymes that are important for ATP generation and that represent distinct protein structural
families. Among phosphoglucomutase (PGM), phosphoglucose isomerase (PGI), pyruvate kinase (PK), phosphoenolpyruvate carboxykinase (GTP) (PEPCK), and isocitrate dehydrogenase (NADP) (IDH), only IDH orthologs showed significantly
different thermal responses of Km between the two species. The Km of isocitrate of M. galloprovincialis-IDH was intrinsically
lower and more thermally stable than that of M. trossulus-IDH and thus had higher substrate affinity at high temperatures.
Two amino acid substitutions account for the functional differences between IDH orthologs, one of which allows for more
hydrogen bonds to form near the mobile region of the active site in M. galloprovincialis-IDH. Taken together, our findings
cast light on the targets of adaptive evolution in the context of climate change; only a minority of proteins might adapt to
small changes in temperature, and these adaptations may involve only small changes in sequence.
MBE
Lockwood and Somero · doi:10.1093/molbev/mss111
Table 1. Enzyme Biochemical and Structural Characteristics (SCOP: Structural Classification of Proteins, 1.75 release, June 2009).
Enzyme
PGM
PGI
PK
PEPCK
IDH
cMDH
Subunit Structure
Monomer
Homodimer
Homotetramer
Monomer
Homodimer
Homodimer
physiology, in terms of cardiac performance, metabolic
rates, enzymatic activity levels, and environmental stress
tolerance, compared with the more cold-adapted M.
trossulus, a native of the North Pacific Ocean (Lockwood
and Somero 2011). At the protein level, the cytosolic paralog of malate dehydrogenase (cMDH) of M. galloprovincialis has more thermally stable structural and functional
properties than the M. trossulus ortholog (Fields et al.
2006), consistent with the adaptive pattern observed in
Rossmann-fold dehydrogenases (Fields and Somero 1998;
Johns and Somero 2004; Dong and Somero 2009). To date,
however, cMDH is the only enzyme for which there has
been a comparative biochemical analysis in these mussels.
Here, we sought to determine the extent to which differential temperature adaptation in M. galloprovincialis and
M. trossulus has influenced the evolution of functional
properties of five enzymes that catalyze steps that are adjacent to cMDH in ATP-generating metabolic pathways,
that is, glycolysis and the TCA cycle. We compared the effects of temperature on the kinetic properties (Km) of orthologs of phosphoglucomutase (PGM; EC: 5.4.2.2),
phosphoglucose isomerase (PGI; EC: 5.3.1.9), pyruvate kinase (PK; EC: 2.7.1.40), phosphoenolpyruvate carboxykinase
(GTP) (PEPCK; EC: 4.1.1.32), and isocitrate dehydrogenase
(NADP) (IDH; EC: 1.1.1.42). We chose enzymes that are
structurally diverse (table 1) (Ward 1977) and for that reason may differ in their sensitivities to temperature. These
enzymes belong to five different structural families of
proteins have distinct types of 3D folds and span a range
of common quaternary structures (table 1). In addition,
these enzymes are experimentally tractable: 1) they are
soluble and straightforward to assay, 2) their biochemistries
are well characterized, and 3) they each have solved
crystal structures to allow for the interpretation of functional divergence in the context of structure–function
relationships.
We report that PGM, PGI, PK, and PEPCK had remarkably similar kinetic properties and thermal sensitivities between orthologs of M. galloprovincialis and M. trossulus,
despite the different thermal optima and tolerance limits
of the species. Conversely, IDH orthologs showed strikingly
different kinetics between the species, likely due to the only
two amino acid substitutions found between orthologs.
These results demonstrate that not all proteins undergo
functional temperature adaptation in response to the magnitude of temperature difference that characterizes the
evolutionary histories of M. trossulus and M. galloprovincialis. Thus, the targets of thermal selection at the protein level
3062
3D Fold
PGM, first three domains
SIS domain
PK beta-barrel domain like
PEP carboxykinase like
Isocitrate/isopropylmalate dehydrogenase like
NAD(P)-binding Rossmann-fold domains
may be limited and depend on the thermal sensitivities of
different protein structural classes. We discuss how structural motifs, notably structural elements found in regions
of the protein that influence conformational changes associated with ligand binding and catalysis, influence temperature adaptation.
Materials and Methods
Sample Collection
We collected mussels from a single population of each
species, M. galloprovincialis from Santa Barbara, CA
(34°24#15$N, 119°41#30$W) and M. trossulus from Newport, OR (44°38#25$N, 124°03#10$W). Thus, the focus of
our study was to discover differences between species
rather than to describe within species variation as might
occur over latitudinal gradients (Logan et al. 2012). Mytilus
galloprovincialis is an alien species in California, having
been introduced from the Mediterranean Sea in the early
20th century (Geller 1999). We chose the sampling locations to be well outside of the hybridization zone of these
two species that exists in central and northern California
(Rawson et al. 1999; Hilbish et al. 2010). Hereby, we sought
to include in our study only individuals from separate species that are known to have divergent genetic backgrounds
(Seed 1992; Rawson et al. 1999; Brannock et al. 2009). Collections at each sampling location were conducted in June
2007, December 2007, and June 2010. Mussels were transported back to Hopkins Marine Station where they were
kept frozen at 70 °C.
Enzyme Preparation
For each species, we conducted partial purification of the
target enzymes in order to obtain preparations that functioned well in the spectrophotometric assay and were stable for periods of weeks. Full purification was deemed
unnecessary because the kinetic properties we examined
tend not to be dependent on the extent of enzyme purification (see Fields et al. 2006). We isolated protein from
a single tissue type, posterior adductor muscle, in order
to avoid the potential presence of multiple isoforms of each
enzyme across different tissue types. PGM and PGI are encoded by genes at single loci (Hall 1985; Pogson 1989), and
while PK, PEPCK, and IDH have two isoforms, bivalves have
been shown to exclusively express single isoforms in muscle
tissue (Mustafa and Hochachka 1971, 1973; Head and
Gabbott 1980). Muscle was dissected from approximately
50 individuals of each species, pooled, and homogenized in
four volumes of ice-cold homogenization buffer (50 mM
Temperature Adaptation in Enzymes of Mytilus · doi:10.1093/molbev/mss111
potassium phosphate, pH 7.0 at 20 °C, 2 mM ethylenediaminetetraacetic acid [EDTA], and 1 mM dithiothreitol)
with an Ultra-Turrax T25 homogenizer (IKA Works,
Staufen, Germany) for three 10 s bursts at 20,500 Hz;
the sample was kept on ice for 1 min between each burst.
The homogenate was centrifuged at 20,000 g for 30 min
at 4 °C. The supernatant was brought to 30% saturation
with ammonium sulfate and stirred on ice for 30 min.
The sample was then centrifuged at 20,000 g for 30
min at 4 °C, the pellet discarded, and the supernatant
brought to 80% saturation with ammonium sulfate. After
stirring on ice for 30 min, the sample was again centrifuged
at 20,000 g for 30 min at 4 °C, the supernatant discarded,
and the pellet resuspended in a minimal volume of homogenization buffer. The sample was again brought to 80% saturation with ammonium sulfate and stored at 4 °C for up
to 6 weeks. This procedure increased the purity and stability of the preparations and preserved greater than 90% of
the activity for each of the five enzymes.
Measurement of Km
We measured the apparent Michaelis–Menten constant
(Km) of the substrates indicated below for each enzyme.
We chose to focus our study on comparing Km values between orthologs for two reasons. First, Km is a measure of
substrate affinity and therefore a clear indicator of enzyme
function (Hochachka and Somero 2002). Second, Km is
highly sensitive to temperature and orthologous enzymes
often differ adaptively in stability of Km values across different ranges of temperature (Fields 2001; Feller 2008); at
normal body temperatures, Km is strongly conserved
among species despite its sensitivity to acute changes in
temperature (Fields and Somero 1998; Fields et al. 2006).
Importantly, temperature-adaptive differences in Km
among orthologs are typically most pronounced at higher
assay temperatures (Hochachka and Somero 2002). For this
reason, that is, to maximize the likelihood of discerning differences between orthologs, we initially surveyed Km at
40 °C; this is a temperature that is acutely lethal for M.
trossulus but only slightly beyond the thermal limits of
M. galloprovincialis (Lockwood and Somero 2011). While
40 °C is a temperature that neither species survives, none
of the enzymes studied was denatured at 40 °C during the
course of the kinetic measurements, as indicated by stability of rates during the assay. For PK, PEPCK, and IDH, we
also measured Km across a range of temperatures (10, 20,
30, and 40 °C) that span most of the thermal ranges of
Mytilus habitats in the North Pacific and the Mediterranean. The temperature was controlled to within
±0.1 °C. All enzymes were analyzed on a Shimadzu 1601
spectrophotometer (Tokyo, Japan) by following changes
in NADH or NADP absorbance at 340 nm. We used
imidazole buffer for the assays in order to preserve the
pH–temperature relationship of the cellular environment
(Yancey and Somero 1978). Mytilus enzyme preparations
and coupling enzymes were desalted with Zeba Spin columns (Thermo Scientific, Rockford, IL) immediately prior
to Km measurement. Km was calculated from the initial
MBE
reaction rates at seven substrate concentrations by a nonlinear least squares regression fit to the Michaelis–Menten
equation using GraphPad Prism version 5.0d for Macintosh
(GraphPad Software, San Diego, CA).
Phosphoglucomutase (PGM; EC: 5.4.2.2)
For PGM, we measured the Km of glucose 1-phosphate
(G1P). The assay conditions were based on Pogson (1989),
with the following modifications: 50 mM imidazole-HCl
(pH 7.0 at 20 °C), 1 mM NADP, 0.004 mM glucose-1,6diphosphate, 2 mM MgCl2, 1 mM EDTA, 5 U/ml glucose6-phosphate dehydrogenase (G6PDH), and 10–500 lM
G1P. The enzyme was incubated in the cuvette with all components of the assay except G1P for 10 min, and the reaction
was started by the addition of G1P.
Phosphoglucose Isomerase (PGI; EC: 5.3.1.9)
For PGI, we measured the Km of fructose 6-phosphate
(F6P). The assay conditions were based on Hall (1985), with
the following modifications: 50 mM imidazole-HCl (pH 7.0
at 20 °C), 1 mM NADP, 8 mM MgCl2, 5 U/ml G6PDH, and
25–500 lM F6P. The reaction was started by the addition of
enzyme.
Pyruvate Kinase (PK; EC: 2.7.1.40)
For PK, we measured the Km of phosphoenolpyruvate (PEP).
The assay conditions were based on Anestis et al. (2007),
with the following modifications: 50 mM imidazole-HCl
(pH 7.5 at 20 °C), 120 mM KCl, 10 mM MgCl2, 4 mM
ADP, 0.2 mM fructose 1,6-bisphosphate, 0.3 mM NADH,
10 U/ml lactate dehydrogenase, and 5–200 lM PEP. The
reaction was started by the addition of enzyme.
Phosphoenolpyruvate Carboxykinase (GTP) (PEPCK; EC:
4.1.1.32)
For PEPCK, we measured the Km of PEP. The assay conditions were based on Harlocker et al. (1991), with the following modifications: 63 mM imidazole-HCl (pH 6.3 at
20 °C), 1 mM MnCl2, 1 mM MgCl2, 20 mM NaHCO3,
1.5 mM IDP, 0.15 mM NADH, 5 U/ml malate dehydrogenase, and 50–1500 lM PEP. All solutions were degassed
prior to assembling the assay cocktail. The enzyme was
added to the cuvette with all components of the assay except for NaHCO3 and the background reaction rate measured before starting the reaction by the addition of
NaHCO3. The initial reaction rate was calculated by subtracting the background rate.
Isocitrate Dehydrogenase (NADP) (IDH; EC: 1.1.1.42)
For IDH, we measured the Km of DL-isocitrate. The assay
conditions were based on Head and Gabbott (1980), with
the following modifications: 100 mM imidazole-HCl (pH 8.2
at 20 °C), 0.15 mM NADP, 4 mM MgCl2, and 1–100 lM DLisocitrate. The reaction was started by the addition of
enzyme.
Sequencing of Idh
To obtain sequences of the 5# and 3# ends of the Idh
cDNAs of each species, we designed primers to a M. galloprovincialis Idh EST sequence (Venier et al. 2009) and
3063
MBE
Lockwood and Somero · doi:10.1093/molbev/mss111
implemented the rapid amplification of cDNA ends (RACE)
protocol (GeneRacer Kit, Invitrogen, Carlsbad, CA). Total
RNA was extracted and purified from the gill tissue of a single individual of each species with a TissueLyser homogenizer (Qiagen, Valencia, CA) and RNeasy kit (Qiagen). RNA
quality was assessed with a NanoDrop spectrophotometer
(NanoDrop Technologies, Wilmington, DE). The Advantage 2 Polymerase Mix (Clontech Laboratories, Inc., Mountain View, CA) was used for the RACE and all subsequent
polymerase chain reaction (PCR) steps. RACE products
were analyzed on 1% agarose gels stained with SYBR Safe
(Invitrogen). Single bands of expected size were excised and
purified on S.N.A.P. columns (Invitrogen) and sent out for
commercial sequencing (Sequetech, Mountain View, CA).
We then used the sequences of the 5# and 3# Idh RACE
products to design primers (Idh full-length forward primer:
CAGCAAGTGCGTGAAAAATG; Idh full-length reverse
primer: TTGCTGCCCCTTATTCAGAC) to amplify and sequence the full-length Idh-coding region from cDNAs of
both species. Total RNA was extracted from the gill tissue
of three M. galloprovincialis and five M. trossulus individuals, as described above, and reverse transcribed into cDNA
with the AffinityScript One-Step RT-PCR kit (Agilent Technologies, Inc., Santa Clara, CA). To amplify Idh, the PCR
conditions were 94 °C for 2 min followed by 35 cycles
of 94 °C for 30 s, 55 °C for 45 s, and 68 °C for 2 min, with
a final extension of 68 °C for 10 min. PCR products were
purified with the Qiaquick PCR purification kit (Qiagen),
cloned into the pCR4-TOPO vector (Invitrogen), transformed into TOP10 chemically competent cells (Invitrogen),
and grown on lysogeny broth agar plates containing 100 lg
ml1 ampicillin. Positive transformants were analyzed via
PCR (T7 forward primer, M13 reverse primer; 94 °C for
3 min, 35 cycles of 94 °C for 30 s, 55 °C for 30 s, and
68 °C for 2 min, final extension of 68 °C for 10 min), cleaned
with ExoSAP-IT (Affymetrix, Santa Clara, CA), and sent for
sequencing (Sequetech). The complete cDNA sequences
and translated products were submitted to GenBank (accession numbers: JQ429379–JQ429390). The molecular
weights and isoelectric points of the protein products were
predicted based on the translated sequences using the ExPASy server (Gasteiger et al. 2005).
There are two isoforms of IDH, which differ in size and
isoelectric point (Henderson 1965); however, previous work
suggests that only one isoform is expressed in Mytilus adductor muscle (Head 1980). We analyzed the IDH activity of
each species on native polyacrylamide gels (PAGE) and isoelectric focusing gels (IEF) in order to 1) verify the presence
of a single IDH isoform in both muscle and gill tissues and 2)
confirm that the cDNA sequences matched the protein for
which we measured Km. Posterior adductor muscle and gill
tissues from a single individual of each species were dissected
and homogenized as described above (see Enzyme Preparation). Native PAGE was run on 12% Tris–HCl Ready Gel precast polyacrylamide gels (Bio-Rad, Hercules, CA), with Tris/
glycine running buffer (25 mM Tris and 192 mM glycine) and
2X native sample buffer (62.5 mM Tris–HCl, pH 8.0, 25%
glycerol, and 0.01% bromophenol blue) at 100 V for 45
3064
min. IEF was run on Read Gel IEF precast polyacrylamide gels
pH 3–10 (Bio-Rad), with 7 mM phosphoric acid (anode
buffer), 20 mM lysine and 20 mM arginine (cathode buffer),
and 2X IEF sample buffer (50% glycerol), at 100 V for 60 min
and 250 V for 120 min. Molecular weight or isoelectric point
standards (Bio-Rad) were loaded in wells adjacent to samples, and both native PAGE and IEF gels were run on the
Mini-PROTEAN system (Bio-Rad). Subsequent to electrophoresis, gels were incubated in 0.2 M Tris–HCl (pH 8.0),
0.25 mM NADP, 0.1 mM nitroblue tetrazolium, 0.3 mM
phenazine methosulfate, 6 mM DL-isocitrate, and 6 mM
MgCl2 for 4 h at 37 °C to stain for IDH activity (Henderson
1965). Gels were then stained in 0.1% coomassie blue, 25%
ethanol, and 8% acetic acid for 30 min and destained in 12%
ethanol and 7% acetic acid to label molecular weight and
isoelectric point standards.
IDH Molecular Modeling
Protein structure templates were queried with the M. trossulus and M. galloprovincialis Idh translated cDNA sequences using the 3D-Jury server (Ginalski et al. 2003).
Homology-based modeling was performed with SwissModel software (Arnold et al. 2006), and protein models
were visualized with DeepView software (Kiefer et al. 2009).
Statistics
Each Km determination was run in triplicate, and data are
presented as means ± standard error of the mean. To determine if there were significant differences in Km between
orthologous enzymes at 40 °C, we compared the means
using a Student’s t-test. To compare the PK, PEPCK, and
IDH orthologs, which were assayed across a range of temperatures, the Km data were linearized with the natural logarithm and slopes and intercepts compared using an
analysis of covariance (ANCOVA; GraphPad Prism). This
analytical procedure supports a rigorous comparative analysis of temperature dependence of Km in different orthologs (Dong and Somero 2009).
Results
Km versus Temperature
When assayed at 40 °C, PGM, PGI, PK, and PEPCK exhibited
Km values that were indistinguishable between orthologous
enzymes of M. galloprovincialis and M. trossulus (fig. 1A–D).
Only IDH showed a significant difference between the species (Student’s t-test, P 5 0.008), with the ortholog of M.
galloprovincialis having a 39% lower Km (i.e., higher binding
affinity) than the ortholog of M. trossulus (fig. 1E). PGM and
PK orthologs had Km values, of G1P and PEP, respectively,
that were relatively low and within the range of physiological concentrations of these substrates (fig. 1A and C)
(Ebberink and de Zwaan 1980), indicating that substrate
binding in these enzymes is relatively stable at 40 °C.
Because M. galloprovincialis and M. trossulus-IDH orthologs had different Km values at 40 °C (fig. 1E), we further
compared IDH Km at 30, 20, and 10 °C (fig. 2C). At lower
temperatures, the Km values were lower, and the absolute
Temperature Adaptation in Enzymes of Mytilus · doi:10.1093/molbev/mss111
MBE
IDH Sequence and Structure
FIG. 1. Michaelis–Menten constants (Km) of orthologs of Mytilus
galloprovincialis (black) and Mytilus trossulus (white) at 40 °C for
(A) PGM, substrate G1P; (B) PGI, substrate F6P; (C) PK, substrate
PEP; (D) PEPCK, substrate PEP; and (E) IDH, substrate DL-isocitrate.
*P 5 0.008, Student’s t-test.
differences between the Km of each species were smaller
(fig. 2C). However, regardless of the assay temperature,
M. galloprovincialis-IDH had a significantly lower Km than
M. trossulus-IDH (fig. 2C; ANCOVA, P , 0.0001). In contrast, neither PK nor PEPCK showed a species-specific temperature versus Km relationship (fig. 2A and B).
The coding regions of M. trossulus and M. galloprovincialis
Idh cDNAs were 1,344 bp long, encoding proteins of 448
amino acids in length (fig. 3). IDH activity was present
as a single band on both native PAGE and IEF gels and
showed identical migration patterns among species and tissue types (data not shown), indicating that all IDH activity
came from a single isoform. The molecular weights (Mw)
and isoelectric points (pI) of the native enzymes matched
the predicted values from the translated sequences: Mw
5 50.5 kDa per monomer and pI 5 6.8 for both orthologs.
There were ten synonymous and two nonsynonymous
fixed differences between the orthologs of the two species,
which resulted in amino acid substitutions at residues 205
and 208 (fig. 3). Site 205 was polymorphic for Asp/Glu in
M. trossulus-IDH but fixed for His in M. galloprovincialis-IDH.
The substitution at residue 205 constituted a flip in charge,
from the negatively charged Asp/Glu-205 side chain in
M. trossulus-IDH to the positive His-205 in M. galloprovincialis-IDH. The substitution at residue 208 changed from the
uncharged Asn-208 in M. trossulus-IDH to the negatively
charged Asp-208 in M. galloprovincialis-IDH (fig. 3). Both
substitutions were in the intermonomer clasp region of
the protein (figs. 3 and 4)(Ceccarelli et al. 2002).
Strikingly, the Asp/Glu-205 to His substitution in M. galloprovincialis-IDH was predicted to allow the formation of
two hydrogen bonds that were not present in M. trossulusIDH (fig. 4B and C). These two additional hydrogen bonds
linked the side chain of His-205 to the side chains of Lys-189
and Glu-191 (fig. 4C). As shown in figure 4C, Lys-189 and
Glu-191 are adjacent to a region of the enzyme that undergoes substantial movement due to conformational changes
during catalysis (Stoddard and Koshland 1993).
Discussion
FIG. 2. Temperature versus Km of orthologs of (A) PK, (B) PEPCK,
and (C) IDH assayed at 10, 20, 30, and 40 °C. Insets show the data
linearized by natural log transformation to facilitate ANCOVA with
corresponding P values.
Despite temperature adaptation being a driver of diversification among orthologous proteins (Fields 2001; Somero
2004), among the five enzymes that we examined, four
were remarkably similar between M. galloprovincialis and
M. trossulus when compared at 40 °C. This is a surprising
result, given the striking differences between orthologs of
cMDH (Fields et al. 2006) and IDH (figs. 1 and 2) in these
species. Previous work has demonstrated functional temperature adaptation in two of these four enzymes in other
systems (PGI in butterflies of the genus Colias [Watt 1977;
Watt et al. 1996; Watt et al. 2003] and PK in fish [Low and
Somero 1976]). In Mytilus edulis on the east coast of the
United States, electrophoretic alleles of PGI exhibit a latitudinal cline (Koehn et al. 1976), suggesting that temperature
may be driving local adaptation at this locus. However,
these allelic variants of PGI show only subtle functional differences related to temperature (Hall 1985). A similar story
holds true for alleles of PGM in the sea anemone Metridium
senile, which exhibit clinal variation on the northeastern
coast of North America but show only subtle differentiation in Km that does not correlate with temperature
(Hoffmann 1985). These observations suggest that certain
3065
Lockwood and Somero · doi:10.1093/molbev/mss111
MBE
FIG. 3. Amino acid sequences of IDHs of Mytilus trossulus (M. t.) and Mytilus galloprovincialis (M. g.) deduced from cDNA sequences. Dashes
indicate conservation with the M. trossulus sequence, and only fixed differences between the species are shown. Substitutions at sites 205 and
208 are highlighted in bold. The IDH intermonomer clasp region is highlighted in black and the highly mobile active site region is highlighted in
red.
enzyme loci may be tightly linked to other regions of the
genome upon which natural selection is acting or that natural selection may act on something other than the Km of
PGM and PGI (Pierce 1982).
The difference in thermal sensitivity of Km between M.
galloprovincialis and M. trossulus-IDH orthologs is consistent with the temperature adaptation seen in dehydrogenase enzymes in other species, including vertebrates and
invertebrates (Fields 2001; Somero 2004). As noted, this
pattern has been observed for cMDH in congeners of Mytilus, including a third species, M. californianus (Fields et al.
2006). Among cMDH orthologs, a single amino acid substitution makes the Km of the M. galloprovincialis ortholog
lower than that of M. trossulus across a range of temperatures (Fields et al. 2006). Similar observations of the differential effects of temperature on Km have been made
among dehydrogenase orthologs of many other ectothermic species, including the cMDHs of limpets of the genus
Lottia (Dong and Somero 2009) and abalone Haliotis
(Dahlhoff and Somero 1993) and A4-lactate dehydrogenases (A4-LDHs) of many fishes and reptiles (Place
and Powers 1979; Graves and Somero 1982; Somero
1995; Fields and Somero 1998; Johns and Somero 2004).
Although M. galloprovincialis-IDH maintains a lower Km
than the ortholog of the native M. trossulus, the two orthologs exhibit similar Km values when assayed at temperatures that reflect the conditions of their respective
native habitats, suggesting that there is a conservation
of IDH substrate-binding ability at physiologically relevant
temperatures. Higher substrate-binding affinity in the ortholog of M. galloprovincialis would be advantageous at
higher temperatures but detrimental to efficient enzyme
product release at lower temperatures (Hochachka and
Somero 2002). Thus, it appears that the IDH ortholog of
M. galloprovincialis is warm adapted, whereas the IDH ortholog of M. trossulus is cold adapted. Although these differences are consistent with the thermal evolutionary
histories of these two species (Thunell 1979; Seed 1992;
Mann and Hamilton 1995), they also correspond to their
present day distributions on the west coast of the North
America (Rawson et al. 1999; Hilbish et al. 2010), suggesting
3066
that adaptation to the thermal environment in the Mediterranean may influence where M. galloprovincialis exerts
its dominance as an invasive species.
The following explanation may help to reconcile the discrepancy among the occurrences of temperature adaptation in the enzymes examined in this study. The differences
between the native thermal habitats of M. galloprovincialis
and M. trossulus, while distinct, are relatively small, on the
order of a few degrees Celsius, when compared with other
studies that have measured temperature adaptation in enzymes from species adapted to much wider ranges of temperature, for example, Antarctic fishes versus temperate
fishes (Fields and Somero 1998) and psychrophilic versus
thermophilic microbes (Bae and Phillips 2004; Siddiqui
and Cavicchioli 2006). The differences in temperature that
characterize the evolutionary histories of the Mytilus congeners may be sufficient to favor selection for temperatureadaptive differences only in classes of proteins that are
especially perturbed by change in temperature. Thus,
we propose that some functional and structural families
of proteins are more sensitive to thermal perturbation than
others and that IDH and cMDH are more sensitive than
PGM, PGI, PK, and PEPCK. This conjecture is supported
to some degree by analysis of genomic data, which have
shown that, even among species whose habitats differ as
much as 100 °C, the degree of divergence among orthologs
varies across the proteome (Gu and Hilser 2009).
A potentially general pattern in temperature adaptation
of enzymes is that functional divergence among orthologs
is dependent on the degree of conformational change that
the enzyme undergoes during catalysis (Fields and Somero
1998; Bae and Phillips 2004), with large movements seeming to require adaptation to maintain function at different
temperatures. Both IDH and cMDH undergo substantial
conformational shifts during catalysis, though the mobile
domains are different in each enzyme (Stoddard and
Koshland 1993). Our data, in addition to those of Fields
et al. (2006), support the idea that regions of proteins
that undergo large conformational changes may be foci
of adaptive change across thermal environments. Specifically, warm-adapted orthologs tend to have amino acid
Temperature Adaptation in Enzymes of Mytilus · doi:10.1093/molbev/mss111
MBE
FIG. 4. (A) Model of the M. trossulus-IDH dimer, based on the crystal structure of pig-IDH (PDB accession: 1LWD). Monomer A is colored in
yellow and monomer B in blue. Amino acid residues are shown for the substitution sites at positions 205 and 208 for both monomers. The
highly mobile region of the active site is highlighted in red. (B) Magnified view of M. trossulus-IDH monomer A, showing residues at positions
205 and 208 in the intermonomer clasp region and mobile region of the active site highlighted in red. (C) M. galloprovincialis-IDH, showing
substituted residues H205 and D208. Dashed lines indicate hydrogen bonding from H205 to K189 and E191 on the adjacent b-sheet of the
intermonomer clasp region. Mobile region of the active site is highlighted in red.
substitutions that stabilize mobile regions (Fields et al.
2006; Dong and Somero 2009), as shown by the additional
hydrogen bonding in M. galloprovincialis-IDH (fig. 4C). The
structure of IDH, characterized by an isocitrate/isopropylmalate dehydrogenase–like fold, is a structural family
closely related to the NAD(P)-binding Rossmann-fold domain of cMDH and A4-LDH (Murzin et al. 1995; Fields and
Somero 1998; Andreeva et al. 2004, 2008). This suggests
that the similar trends in temperature adaptation for
IDH and cMDH in Mytilus (Fields et al. 2006) and for
A4-LDHs of differently adapted vertebrates (Fields and
Somero 1998) are due to the shared aspects of their 3D
structures. Thus, the structural motifs present in these enzymes that undergo large conformational changes during
ligand binding may establish a high thermal sensitivity of
function in these proteins.
A less likely explanation for the lack of functional divergence in Mytilus PGM, PGI, PK, and PEPCK is that natural
selection may constrain the evolution of enzymes that
function either upstream in biochemical pathways
(i.e., PGM and PGI) (Rausher et al. 1999) or enzymes that
catalyze rate-limiting steps (i.e., PK and PEPCK) (Hartl et al.
1985). While PGM and PGI function upstream of glycolysis,
they are downstream of gluconeogenesis; thus, it is unclear
3067
MBE
Lockwood and Somero · doi:10.1093/molbev/mss111
whether natural selection would constrain PGM and PGI
when they function in juxtaposed pathways. The evolutionary constraint hypothesis also does not hold for the
rate-limiting enzymes that we examined, as IDH catalyzes
the rate-limiting step in the TCA cycle, and it is the enzyme
for which we observed the greatest divergence in Km between M. galloprovincialis and M. trossulus. Furthermore,
PK orthologs of vertebrates from widely different thermal
environments exhibit the type of Km patterns noted in
studies of IDH, cMDH, and LDH (Somero and Hochachka
1968). Therefore, selection that constrains the evolution of
rate-limiting enzymes in other contexts might also cause
adaptive change that preserves function in a different thermal environment.
Given that we compared Km responses with temperature between orthologs of only five enzymes, our data provide but a preliminary snapshot of the breadth of
occurrence of temperature adaptation at the protein level.
Comparative genomics studies give a more comprehensive
view of the proteome (Gu and Hilser 2009; Oliver et al.
2010) but do not directly measure function as we have
done here. This is an important distinction between
approaches because while there have been important
insights about temperature adaptation gained from in
silico methods (Goldstein 2011), these studies have been
aimed at describing patterns of overall ‘‘global’’ protein
stability rather than maintenance of normal protein function (Gu and Hilser 2009). The distinction between global
denaturation—wholesale loss of protein higher order
structure—and subtle thermal disruption of protein structure that falls short of complete inactivation of the protein
must be taken into account in all such analyses. Perturbations to protein stability that lead to denaturation occur at
more extreme temperatures than thermally induced
changes in functionally important properties like ligand
binding, as indexed by Km (Fields and Somero 1998; Fields
et al. 2006; Feller 2008). Furthermore, adaptation of function through adjustments in local-scale stability of mobile
regions can occur without a concomitant increase (or decrease) in global protein stability (Fields and Somero 1998;
Arnold et al. 2001; Daniel et al. 2003, 2008; Bae and Phillips
2006). It follows that amino acid substitutions that modify
global stability may arise only when differences in thermal
evolutionary history reach levels that are substantially
larger than the temperature changes that are sufficient
to select for temperature-adaptive variation in traits like
Km. In silico studies have, to our knowledge, not attempted
to differentiate between amino acid substitutions linked to
changes in global stability versus those that ‘‘fine-tune’’
the stabilities of localized regions in a protein that govern
binding events.
Future study should integrate in vitro and in silico approaches as complementary methods. Through an in vitro
comparison of enzyme function, we have demonstrated
a degree of functional adaptation to temperature in Mytilus
that provides a predictive framework for protein evolution
in ectotherms in the context of climate change (Somero
2010). As sequence data in Mytilus become more widely
3068
available, it will be possible to assess the degree of divergence among a broader array of proteins and thereby generate hypotheses about the targets of thermal selection
that can be tested by in vitro functional assays. Such an
approach would elucidate the structural characteristics,
for example, occurrence of different structural motifs
and extents of conformational changes during function
that distinguish proteins with thermal sensitivities that
are predisposed to temperature adaptation.
Acknowledgments
We thank Dafne Eerkes-Medrano, Jennie Yoder, Rachel
Mahler, Elizabeth Lopez, Maja Haji, and Alex Hirota for collecting Mytilus trossulus. We thank Dr Ward Watt for helpful suggestions on the protein structural modeling analysis.
We also thank Dr Kristi Montooth, Dr Melissa Pespeni,
Brandon Cooper, Luke Hoekstra, Robert Kobey, Mo Siddiq,
and two anonymous reviewers for their helpful comments
on this manuscript. This work was funded by National Science Foundation (NSF) Graduate Research Fellowship to
B.L.L., NSF grant IOS-0718734 to G.N.S., the Earl and Ethel
Myers Marine Biological and Oceanographic Trust to B.L.L.,
the Jane Miller Scholars Award to B.L.L., the Friends of Hopkins Scholar Award to B.L.L., and the Partnership for the
Interdisciplinary Study of the Coastal Oceans (PISCO)
sponsored by the David and Lucile Packard Foundation
and Gordon and Betty Moore Foundation. This is PISCO
publication number 416.
References
Andreeva A, Howorth D, Brenner SE, Hubbard TJP, Chothia C,
Murzin AG. 2004. SCOP database in 2004: refinements
integrate structure and sequence family data. Nucleic Acids
Res. 32:D226.
Andreeva A, Howorth D, Chandonia JM, Brenner SE, Hubbard TJP,
Chothia C, Murzin AG. 2008. Data growth and its impact on
the SCOP database: new developments. Nucleic Acids Res.
36:D419.
Anestis A, Lazou A, Pörtner HO, Michaelidis B. 2007. Behavioral,
metabolic, and molecular stress responses of marine bivalve
Mytilus galloprovincialis during long-term acclimation at increasing ambient temperature. Am J Physiol Regul Integr Comp
Physiol. 293:R911–R921.
Arnold FH, Wintrode PL, Miyazaki K, Gershenson A. 2001. How
enzymes adapt: lessons from directed evolution. Trends Biochem
Sci. 26:100–106.
Arnold K, Bordoli L, Kopp J, Schwede T. 2006. The swiss-model
workspace: a web-based environment for protein structure
homology modelling. Bioinformatics 22:195–201.
Bae E, Phillips GN. 2004. Structures and analysis of highly
homologous psychrophilic, mesophilic, and thermophilic adenylate kinases. J Biol Chem. 279:28202–28208.
Bae E, Phillips GN. 2006. Roles of static and dynamic domains in
stability and catalysis of adenylate kinase. Proc Natl Acad Sci U S A.
103:2132–2137.
Berezovsky IN, Shakhnovich EI. 2005. Physics and evolution of
thermophilic adaptation. Proc Natl Acad Sci U S A.
102:12742–12747.
Braby CE, Somero GN. 2006. Following the heart: temperature and
salinity effects on heart rate in native and invasive species of
blue mussels (genus Mytilus). J Exp Biol. 209:2554–2566.
Temperature Adaptation in Enzymes of Mytilus · doi:10.1093/molbev/mss111
Brannock P, Wethey D, Hilbish T. 2009. Extensive hybridization
with minimal introgression in Mytilus galloprovincialis and
M. trossulus in Hokkaido, Japan. Mar Ecol Prog Ser. 383:161–171.
Ceccarelli C, Grodsky NB, Ariyaratne N, Colman RF, Bahnson BJ.
2002. Crystal structure of porcine mitochondrial NADPþdependent isocitrate dehydrogenase complexed with Mn2þ
and isocitrate. J Biol Chem. 277:43454–43462.
Dahlhoff EP, Somero GN. 1993. Kinetic and structural adaptations of
cytoplasmic malate dehydrogenases of eastern Pacific abalone
(genus Haliotis) from different thermal habitats: biochemical
correlates of biogeographical patterning. J Exp Biol. 185:137–150.
Daniel RM, Danson MJ, Hough DW, Lee CK, Peterson ME,
Cowan DA. 2008. Enzyme stability and activity at high temperatures. In: Siddiqui KS, Thomas T, editors. Protein adaptation in
extremophiles. New York: Nova Science Publishers, Inc. p. 1–34.
Daniel RM, Dunn RV, Finney JL, Smith JC. 2003. The role of dynamics
in enzyme activity. Annu Rev Biophys Biomol Struct. 32:69–92.
Dong YW, Somero GN. 2009. Temperature adaptation of cytosolic
malate dehydrogenases of limpets (genus Lottia): differences in
stability and function due to minor changes in sequence
correlate with biogeographic and vertical distributions. J Exp
Biol. 212:169–177.
Ebberink RHM, de Zwaan A. 1980. Control of glycolysis in the
posterior adductor muscle of the sea mussel Mytilus edulis.
J Comp Physiol B. 137:165–171.
Feller G. 2008. Enzyme function at low temperatures in psychrophiles. In: Siddiqui KS, Thomas T, editors. Protein adaptation in
extremophiles. New York: Nova Science Publishers, Inc. p. 35–69.
Fields PA. 2001. Review: protein function at thermal extremes:
balancing stability and flexibility. Comp Biochem Physiol A Mol
Integr Physiol. 129:417–431.
Fields PA, Kim Y-S, Carpenter JF, Somero GN. 2002. Temperature
adaptation in Gillichthys (Teleost: Gobiidae) A4-lactate dehydrogenases: identical primary structures produce subtly different
conformations. J Exp Biol. 205:1293–1303.
Fields PA, Rudomin EL, Somero GN. 2006. Temperature sensitivities
of cytosolic malate dehydrogenases from native and invasive
species of marine mussels (genus Mytilus): sequence-function
linkages and correlations with biogeographic distribution. J Exp
Biol. 209:656–667.
Fields PA, Somero GN. 1998. Hot spots in cold adaptation: localized
increases in conformational flexibility in lactate dehydrogenase
A4 orthologs of Antarctic notothenoid fishes. Proc Natl Acad Sci
U S A. 95:11476–11481.
Gasteiger E, Hoogland C, Gattiker A, Duvaud S, Wilkins MR,
Appel RD, Bairoch A. 2005. Protein identification and analysis
tools on the ExPASy server. In: Walker JM, editor. The
proteomics protocols handbook. Totowa (NJ): Humana Press.
p. 571–607.
Geller JB. 1999. Decline of a native mussel masked by sibling species
invasion. Conserv Biol. 13:661–664.
Ginalski K, Elofsson A, Fischer D, Rychlewski L. 2003. 3D-jury:
a simple approach to improve protein structure predictions.
Bioinformatics 19:1015–1018.
Goldstein RA. 2011. The evolution and evolutionary consequences
of marginal thermostability in proteins. Proteins Struct Funct
Bioinform. 79:doi:10.1002/prot.22964
Graves JE, Somero GN. 1982. Electrophoretic and functional
enzymatic evolution in 4 species of eastern Pacific barracudas
from different thermal environments. Evolution 36:97–106.
Gu J, Hilser VJ. 2009. Sequence-based analysis of protein energy
landscapes reveals nonuniform thermal adaptation within the
proteome. Mol Biol Evol. 26:2217–2227.
Hall JG. 1985. Temperature-related kinetic differentiation of
glucosephosphate isomerase alleloenzymes isolated from the
blue mussel, Mytilus edulis. Biochem Genet. 23:705–728.
MBE
Harlocker SL, Kapper M, Greenwalt DE, Bishop SH. 1991. Phosphoenolpyruvate carboxykinase from ribbed mussel gill tissue: reactivity with metal ions, kinetics, and action of 3-mercaptopicolinic
acid. J Exp Zool. 257:285–298.
Hartl DL, Dykhuizen DE, Dean AM. 1985. Limits of adaptation: the
evolution of selective neutrality. Genetics 111:655–674.
Head EJH. 1980. NADP-dependent isocitrate dehydrogenase from
the mussel Mytilus edulis L. 1. Purification and characterisation.
Eur J Biochem. 111:575–579.
Head EJH, Gabbott PA. 1980. Properties of NADP-dependent
isocitrate dehydrogenase from the mussel Mytilus edulis L.
Comp Biochem Physiol B. 66:285–289.
Henderson NS. 1965. Isozymes of isocitrate dehydrogenase: subunit
structure and intracellular location. J Exp Zool. 158:263–273.
Hilbish T, Brannock P, Jones K, Smith A, Bullock B, Wethey D. 2010.
Historical changes in the distributions of invasive and endemic
marine invertebrates are contrary to global warming predictions:
the effects of decadal climate oscillations. J Biogeogr. 37:423–431.
Hochachka PW, Somero GN. 2002. Biochemical adaptation:
mechanism and process in physiological evolution. New York:
Oxford University Press.
Hoffmann RJ. 1985. Properties of allelic variants of phosphoglucomutase from the sea anemone Metridium senile. Biochem Genet.
23:859–876.
Jaenicke R. 1991. Protein stability and molecular adaptation to
extreme conditions. Eur J Biochem. 202:715–728.
Johns GC, Somero GN. 2004. Evolutionary convergence in adaptation
of proteins to temperature: A4-lactate dehydrogenases of Pacific
damselfishes (Chromis spp.). Mol Biol Evol. 21:314–320.
Kiefer F, Arnold K, Künzli M, Bordoli L, Schwede T. 2009. The Swissmodel repository and associated resources. Nucleic Acids Res.
37:D387–D392.
Koehn RK, Milkman R, Mitton JB. 1976. Population genetics of marine
pelecypods. IV. Selection, migration and genetic differentiation in
the blue mussel Mytilus edulis. Evolution 30:2–32.
Lockwood BL, Sanders JG, Somero GN. 2010. Transcriptomic
responses to heat-stress in invasive and native blue mussels
(genus Mytilus): molecular correlates of invasive success. J Exp
Biol. 213:3548–3558.
Lockwood BL, Somero GN. 2011. Invasive and native blue mussels
(genus Mytilus) on the California coast: the role of physiology in
a biological invasion. J Exp Mar Biol Ecol. 400:167–174.
Logan CA, Kost LE, Somero GN. 2012. Latitudinal differences in
Mytilus californianus thermal physiology. Mar Ecol Prog Ser.
450:93–105.
Low PS, Somero GN. 1976. Adaptation of muscle pyruvate kinases to
environmental temperatures and pressures. J Exp Zool. 198:1–11.
Mann D, Hamilton T. 1995. Late Pleistocene and Holocene paleoenvironments of the North Pacific coast. Q Sci Rev. 14:449–471.
Murzin AG, Brenner SE, Hubbard T, Chothia C. 1995. SCOP:
a structural classification of proteins database for the investigation of sequences and structures. J Mol Biol. 247:536–540.
Mustafa T, Hochachka PW. 1971. Catalytic and regulatory properties
of pyruvate kinases in tissues of a marine bivalve. J Biol Chem.
246:3196–3203.
Mustafa T, Hochachka PW. 1973. Enzymes in facultative anaerobiosis of molluscs—II. Basic catalytic properties of phosphoenolpyruvate carboxykinase in oyster adductor muscle. Comp
Biochem Physiol B. 45:639–655.
Oliver TA, Garfield DA, Manier MK, Haygood R, Wray GA,
Palumbi SR. 2010. Whole-genome positive selection and
habitat-driven evolution in a shallow and a deep-sea urchin.
Genome Biol Evol. 2:800–814.
Pierce S. 1982. Invertebrate cell volume control mechanisms:
a coordinated use of intracellular amino acids and inorganic
ions as osmotic solute. Biol Bull. 163:405–419.
3069
Lockwood and Somero · doi:10.1093/molbev/mss111
Place AR, Powers DA. 1979. Genetic variation and relative catalytic
efficiencies—lactate dehydrogenase-b allozymes of Fundulus
heteroclitus. Proc Natl Acad Sci U S A. 76:2354–2358.
Pogson GH. 1989. Biochemical characterization of genotypes at the
phosphoglucomutase-2 locus in the Pacific oyster, Crassostrea
gigas. Biochem Genet. 27:571–589.
Rausher MD, Miller RE, Tiffin P. 1999. Patterns of evolutionary rate
variation among genes of the anthocyanin biosynthetic
pathway. Mol Biol Evol. 16:266–274.
Rawson P, Agrawal V, Hilbish T. 1999. Hybridization between the
blue mussels Mytilus galloprovincialis and M. trossulus along the
Pacific Coast of North America: evidence for limited introgression. Mar Biol. 134:201–211.
Seed R. 1992. Systematics, evolution and distribution of mussels
belonging to the genus Mytilus: an overview. Am Malacol Bull.
9:123–137.
Siddiqui KS, Cavicchioli R. 2006. Cold-adapted enzymes. Annu Rev
Biochem. 75:403–433.
Somero GN. 1995. Proteins and temperature. Annu Rev Physiol.
57:43–68.
Somero GN. 2004. Adaptation of enzymes to temperature: searching
for basic ‘‘strategies.’’ Comp Biochem Physiol B. 139:321–333.
Somero GN. 2010. The physiology of climate change: how potentials
for acclimatization and genetic adaptation will determine
‘winners’ and ‘losers.’ J Exp Biol. 213:912–920.
Somero GN, Hochachka PW. 1968. The effect of temperature on
catalytic and regulatory functions of pyruvate kinases of the
rainbow trout and the Antarctic fish Trematomus bernacchii.
Biochem J. 110:395–400.
Stoddard BL, Koshland DE Jr. 1993. Structure of isocitrate
dehydrogenase with alpha-ketoglutarate at 2.7-A resolution:
3070
MBE
conformational changes induced by decarboxylation of isocitrate. Biochemistry 32:9317–9322.
Thunell R. 1979. Pliocene-pleistocene paleotemperature and
paleosalinity history of the Mediterranean sea: results from
DSDP sites 125 and 132. Mar Micropaleontol. 4:173–187.
Tomanek L, Zuzow MJ. 2010. The proteomic response of the mussel
congeners Mytilus galloprovincialis and M. trossulus to acute
heat stress: implications for thermal tolerance limits and
metabolic costs of thermal stress. J Exp Biol. 213:3559–3574.
Venier P, De Pitta C, Bernante F, et al. (11 co-authors). 2009.
Mytibase: a knowledgebase of mussel (M. galloprovincialis)
transcribed sequences. BMC Genomics 10:72.
Vermeij G. 1991. Anatomy of an invasion: the trans-arctic
interchange. Paleobiology 17:281–307.
Ward RD. 1977. Relationship between enzyme heterozygosity and
quaternary structure. Biochem Genet. 15:123–135.
Watt WB. 1977. Adaptation at specific loci. I. Natural selection on
phosphoglucose isomerase of Colias butterflies: biochemical and
population aspects. Genetics 87:177–194.
Watt WB, Donohue K, Carter PA. 1996. Adaptation at specific loci.
VI. Divergence vs. parallelism of polymorphic allozymes in
molecular function and fitness-component effects among Colias
species (Lepidoptera, Pieridae). Mol Biol Evol. 13:699–709.
Watt WB, Wheat CW, Meyer EH, Martin JF. 2003. Adaptation at
specific loci. VII. Natural selection, dispersal and the diversity of
molecular-functional variation patterns among butterfly species
complexes (Colias: Lepidoptera, Pieridae). Mol Ecol. 12:1265–1275.
Yancey PH, Somero GN. 1978. Temperature dependence of
intracellular pH: its role in the conservation of pyruvate
apparent Km values of vertebrate lactate dehydrogenases.
J Comp Physiol B. 125:129–134.