Seasonal variations of gene expression biomarkers in Mytilus galloprovincialis
cultured populations: temperature, oxidative stress and reproductive cycle as
major modulators
Sergio Jarque1,2,*, Eva Prats1, Alba Olivares1, Marta Casado1, Montserrat Ramón3,4,
Benjamin Piña1
1)
Institute of Environmental Assessment and Water Research (IDAEA-CSIC). Jordi
Girona, 18. 08034 Barcelona, Spain.
2)
Masaryk University, Faculty of Science, RECETOX. Kamenice 753/5, 625 00, Brno,
Czech Republic.
3)
IEO-Centre Oceanogràfic de les Balears, Moll de Ponent s/n, 07015 Palma, Spain
4)
Institut de Ciències del Mar (CSIC), Passeig Marítim de la Barceloneta, 37-49, 08003
Barcelona, Spain
*Corresponding author. Tel: +420 549 49 5305 e-mail: jarque@recetox.muni.cz
ABSTRACT
The blue mussel Mytilus galloprovincialis has been used as monitoring organism in
many biomonitoring programs because of its broad distribution in South European sea
waters and its physiological characteristics. Different pollution-stress biomarkers,
including gene expression biomarkers have been developed to determine its
physiological response to the presence of different pollutants. However, the existing
information about basal expression profiles is very limited, as very few biomarkerbased studies were designed to reflect the natural seasonal variations. In the present
study, we analyzed the natural expression patterns of several genes commonly used in
biomonitoring, namely ferritin, metallothionein, cytochrome P450, glutathione Stransferase, heat shock protein and the kinase responsive to stress KRS, during an
annual life cycle. Analysis of mantle-gonad samples of cultured populations of M.
galloprovincialis from the Delta del Ebro (North East Spain) showed natural seasonal
variability of these biomarkers, pointing to temperature and oxidative stress as major
abiotic modulators. In turn, the reproductive cycle, a process that can be tracked by
VCLM7 expression, and known to be influenced by temperature, seems to be the major
biotic factor involved in seasonality. Our results illustrate the influence of
environmental factors in the physiology of mussels through their annual cycle, a crucial
information for the correct interpretation of responses under stress conditions.
Keywords: Mytilus galloprovincialis, qRT-PCR, seasonal variations, gene expression,
biomonitoring
Abbreviations: actin, β-actin; AU: arbitrary units; bp; base pair; CYP, cytochrome
P450 family members; DO, dissolved oxygen; Fer, Ferritin; KRS, kinase responsive to
stress; GSTpi1, glutathione S-trasnferase pi1; HKG, housekeeping gene; HSE, heat
shock element; HSP70, heat shock protein 70; I, immature; L13A; ribosomal protein
L13A; L19, ribosomal protein L19a; MT, metallothionein family members; PAHs,
polycyclic aromatic hydrocarbons; PCBs, polychlorinated biphenyls; PC1,2,3, principal
component 1,2,3; Pre-S, pre-spawning; Post-S, post-spawning; S3, ribosomal protein
S3; Temp, temperature; TG, tested gene; VCLM7, viteline coat lysine M7.
1. Introduction
Mussels of the genus Mytilus are among the commonest marine mollusks and
constitute an important element in the ecology of coastal waters. Three taxa or forms of
the genus Mytilus inhabits along the European coast, two of them being predominant:
M. edulis, which occupies temperate to cold areas along European Atlantic coasts, and
M. galloprovincialis, a warm-water form that occurs in the Mediterranean and extends
northward to the coast of France and the United Kingdom. Apart from their ecological
importance and economic value, mussel species have gained an important role as
bioindicators because they are sessile and filter feeders, which results in the
accumulation of contaminants in their tissues. They also provide the opportunity of
allowing comparison between cultivated and wild populations, making them especially
interesting for biomonitoring (Marigómez et al., 2013; Serafim et al., 2011). As
consequence, mussels have been commonly used as monitoring organisms in several
environmental studies, and a battery of pollution-stress biomarkers have been
developed, including biochemical, cytological, genetic and gene expression-based
assays (Tanguy et al., 2008; Porte et al., 2006; Saavedra and Bachere, 2006; Venier et
al., 2006, 2003).
Among the gene biomarkers used in biomonitoring are those relating to
oxidative stress and metal and organic contamination. However, although some of them
have been successfully applied in several environmental studies (Serafim et al., 2011;
Bebianno et al., 2007), there are still some limitations that may compromise their
validity when using mussels as monitoring organisms (Forbes et al., 2006). One of the
major drawbacks refers to the current limited knowledge of the invertebrate physiology,
which makes difficult to distinguish between natural physiological responses from those
considered as stressful. In addition, the fact that mussels colonize intertidal
environments influenced by daily and annual cycles may increase natural variability and
hinder interpretation of results (Gracey et al., 2008). So far, most of the environmental
studies using gene biomarkers have been focused on determining causation between
exposure to pollutants and changes in gene expression, and only very few have
considered natural seasonal variations as gene expression modulators (Schmidt et al.,
2013; Banni et al., 2011). This lack of information is particularly relevant since a
number of genes have been demonstrated to respond to several biotic and abiotic factors
(Schmidt et al., 2013; Banni et al., 2011; Luedeking et al., 2004). In this respect, the
study of populations located in unpolluted areas is postulated as a mandatory task to
fully understand the natural gene expression profiles and, in turn, to interpret correctly
responses under stress conditions.
In the present study, we analyzed the natural expression patterns of several gene
biomarkers during the course of an annual life cycle in the mantle-gonad of cultured
populations of M. galloprovincialis; β-actin (actin) as structural protein often used as
reference gene, viteline coat lysine M7 (VCLM7) as male-specific maturation stage
biomarker, ferritin (Fer) as indicator for anoxia, metallothionein (MT-10) for metal
contamination, cytochrome P450 (CYP4Y1) and glutathione S-transferase (GSTpi1) for
oxidative stress, and heat shock protein (HSP70) and kinase responsive to stress (KRS)
as general stress biomakers. Because the study was carried out with samples from Ebro
Delta, a well-known relatively unpolluted area (Sole et al., 2000; 1994), the observed
seasonal variations are considered to be presumably associated to natural expression
patterns.
2. Materials and methods
2.1 Physical and chemical data.
Temperature data at Alfacs bay (Ebro Delta, NW Mediterranean Sea) at the time
of mussel collection was obtained from an HOBO Water Temperature Pro v2 Data
Logger (ONSET Computer Corporation), deployed at the mussel farm. Missing values
(two samplings, in may and june 2006) were parametrized using the corresponding data
from the Ebro River at the Tortosa station, just at the beginning of the Delta, data from
the Confederación Hidrografica del Ebro, CHE, www.chebro.es). Changes on the
salinity levels at the Delta bays were calculated from the historical record (1990-2004,
Llebot et al., 2011)
2.2 Mussel sampling
A commercial culture of Mytilus galloprovincialis from Alfacs bay (Ebro Delta)
was monitored from March 2005 to July 2006 in order to study growth, mortality and
reproduction. To analyze the expression of the genes, four to six individuals were
sampled monthly (64 mussels in total) and their sex (male/female) and maturation stage
(mature/immature) checked from gonadal smears under a dissection microscope. Total
length and weight of the specimens were measured. A fragment of mantle tissue (about
100 mg), which includes the gonad when the mussel is mature, was dissected, frozen in
liquid nitrogen and stored at -80ºC until processing.
2.3. RNA preparation and qRT-PCR analysis
Total RNA was extracted from individual frozen mantles using Trizol (SigmaAldrich, Buchs SG, Switzerland) as previously described (Garcia-Reyero et al., 2004).
Total RNA concentration was estimated by spectrophotometric absorption at 260 nm in
a NanoDrop ND-1000 Spectrophotometer (NanoDrop Technologies; Delaware, DE).
RNA integrity was checked in a Bioanalyzer (Agilent Technologies Inc., Palo Alto, CA,
USA). One to five µg of RNA per sample was afterwards treated with DNase I (F.
Hoffmann-La Roche Ltd., Basel, Switzerland), retro-transcribed to cDNA (Omniscript,
Qiagen, Valencia, CA) and stored at -20ºC. Specific transcripts were quantified by Real
Time PCR in a Abi Prism 7000 SDS (Applied Biosystems, Foster City, CA) using the
SYBR Green chemistry (Power SYBR Green PCR Master Mix, Applied Biosystems).
Primer sequences and Gene Bank references are detailed in Table 1. Preliminary results
showed that the variability among individuals may act as confounding factor in some
cases. Consequently, samples were considered individually rather than in pools based
on month or maturation status.
2.3. Data analysis and statistics
Relative expression values were calculated according to equation 1 using
threshold cycle (Ct) values from triplicate assays as previously described (Pfaffl, 2001):
Ct HGK 
mRNA TG
 EHGKCt TG   1000
ETG
mRNA HGK
Equation 1
In which TG and HKG indicate tested and reference genes, respectively. Evaluation of
the suitability of different reference genes was tested by the BestKeeper programm
(Pfaffl et al., 2004). PCR efficiency values for reference and tested genes, EHGK and
ETG, were calculated as described (Pfaffl, 2001), both of them being closed to100%. The
sequence of amplified PCR products (amplicons) was confirmed by DNA sequencing in
Applied Biosystems 3730 DNA Analyzer (Applied Biosystems). Amplified sequences
were compared to the corresponding references in GenBank (Table 1) by the BLAST
algorithm at NCBI server (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi).
All statistics, including Partial Correlations and Principal Component Analysis
or PCA were performed using the SPSS 19 (SPSS Inc, 2002) package. The rationale of
using PCA was to group the data set into few variables that could explain most of the
variance associated to the samples. Data from qRT-PCR were analyzed using ∆Cp
values (Cp reference –Cp target), as this parameter follows a normal distribution,
assessed by Kolmogorov –Smirnov test. Statistical comparison of mean values was
done using One Way analysis of variance (ANOVA) plus Tukey's tests. Correlograms
were drawn using the corrgram v.15 library in R.
3. Results
3.1. Determination of reference genes for qRT-PCR analysis in the developing mussel
gonad.
To determine the suitability of different candidates for reference genes in
Mytilus, we analyzed Ct values for ß-actin and the three ribosomal proteins S3, L13A
and L19 in16 individuals (4 males, 4 females and 8 immature) captured in March, July,
August and December, to roughly cover a complete year. Statistical analysis (Best
Keeper) showed essentially invariant expression levels for all three ribosomal protein
genes considered (standard deviation lower than 1, the cut-off value established by
BestKeeper program). In contrast, ß-actin mRNA levels changed up to 1.7 fold,
showing both individual and seasonal variations. As these differences correlated with
temperature and maturation stage, ß-actin may be considered a biomarker rather than a
housekeeping gene (Fig. 1A, B).
3.2. seasonal variations on sex-specific VCLM7.
Expression of VCLM7 gene showed four to five orders of magnitude variations
among individuals. Low expression rates corresponded to females, immature animals or
males collected from May to July, whereas maximal values occurred in males sampled
from December to April (Fig. 2A). From May to December, there was a progressive
differentiation of the animals into low- and high VCLM7-producers.
3.3. Seasonal variations on stress gene biomarkers.
All genes showed seasonal variations, and correlated with temperature in
greater or lesser degree, except in the case of KRS. Additional masked correlations were
found after suppressing the predominant role of temperature as variable. Table 2
summarizes correlations between gene expression values and temperature.
Given that one of the main aims in the present work was to try to find common
gene expression patterns related to natural variables, we opted to look for correlations
between these genes, rather than testing individual gene expressions. PCA defined three
principal components (PC1, PC2 and PC3) that explained 70% of the variability in gene
expression (37% for PC1, 18% for PC2 and 15% for PC3, Figure 3A). PC1 showed a
strong contribution from MT-10, Fer, actin and KRS genes, all of them highly
influenced by temperature. PC2 and PC3 mainly correlated to CYP4Y1 and Fer, and
GSTpi1, respectively. Score plots for PC1 and PC2 reveals that the combination of both
components separates males and females pre-spawning from the post-spawning and
immature individuals, revealing a strong effect of reproduction cycle on the expression
of the corresponding gene biomarkers (Figure 3B). Effect of reproduction cycle was
further individually analyzed, and all genes, with the only exceptions of KRS and
CYP4Y1, showed significant differences depending on the maturation stage (Fig. 1B;
Fig. 4).
Correlations between gene expression data and different physical and
chemical parameters were analyzed by Pearson correlation (see the correlograms
in Figure 5, the complete analysis is shown as supplementary material, Table S1).
Most parameters were strongly dependent on temperature (Tables 2 and S1),
either positively (size and weight and most of gene expression data) or negatively
(HSP70, CCLM and GSTpi1, Table 2, Figure 5 Top). When the influence of
temperature (and, likely, seasonal variations) was controlled (Partial correlations),
two clusters of variables become evident (Table S1, Figure 5, bottom). The largest
cluster corresponded to genes that could be considered as "resistance genes" (MT10, Fer, GSTpi1, KRS). The expression of these genes was negatively correlated
with salinity levels and positively correlated with the size of the animals (Table S1,
Figure 5, bottom). The two remaining gene expression datasets, HSP70 and
VCLM7, appear negatively correlated with several of the "resistance genes" group,
both in the original Pearson correlations or in the partial correlation analysis
(Table S1, Figure 5).
4. Discussion
The common mussel has been widely used in environmental studies as bioindicator organism (some examples are Sureda et al., 2011; Banni et al., 2007; Alabat
et al., 2002; Orbea et al., 2002). Authors have used an extensive number of biomarkers
to try to elucidate the effects associated to chemical exposure. However, unless other
biomarkers, relatively few studies have considered natural intra-species seasonal
variations as a gene biomarker modulation factor, so information about basal expression
profiles is very limited. As consequence, the information obtained from biomonitoring
studies may be biased (Cravo et al., 2009), which may lead to erroneous conclusions.
Certainly, results show that most of the genes evaluated in our study naturally vary
during the annual cycle of M. galloprovincialis depending on seasonality.
4.1. Determination of reference gene: ribosomal proteins vs β-actin
A major difficulty on using qRT-PCR techniques to monitor changes on
expression levels of specific transcripts is to identify a gene whose expression does not
change under the conditions of the experiment (Piña et al., 2007; Thellin et al., 1999).
This problem is especially severe when the tissue being analyzed is in a developing or
degeneration process, as it occurs in the mussel mantle-gonad during the reproductive
cycle. Our analysis using BestKeeper program showed that expressions of ribosomal
proteins (S3, L13A and L19) are particularly stable and, therefore, convenient reference
genes for qRT-PCR analyses in gonad of Mytilus. S3, thus, was used to normalize qRTPCR expression values in all samples referred in this study. By contrast, actin, one of
the genes more commonly used as an endogenous reference in biomonitoring so far
(Woo et al., 2011), showed both high individual and seasonal variability, with strong
correlation with temperature (Fig. 1A). In a recent study, the suitability of using actin as
reference gene had already being questioned, since its expression appears to be highly
influenced by the stage of gametogenesis (Cubero-Leon et al., 2011). Further, our
results showed that actin expression (Fig. 1B) was highly enhanced during gonadal
inactivity (June-September) and gradually diminished from gonad inactivity to
gametogenesis and ripeness. This expression pattern is inversely related with the
maturation stage (Fig 2B) and, in our opinion, may point out actin as an important
structural protein within the process, likely involved in tasks of tissue regeneration.
4.2. VCLM7 as a male-specific gene and potentially suitable endocrine disruption
biomarker
There is a growing interest to understand the role of steroids in sex maturation in
bivalves, as well as to determine the mechanisms by which this process is affected by
natural hormones and exogenous endocrine disruptors (Ciocan et al., 2011). So far,
many of the efforts have been focused to feminization effects since most of the known
sex-related biomarkers in vertebrates are related to the estradiol response (Ciocan et al.,
2010; Ortiz-Zarragoitia and Cajaraville, 2006, Puinean and Rotchell, 2006), while malespecific response have attracted less attention. Moreover, expression patterns across full
reproductive cycle have been usually not considered, and only very recently, some
groups have pointed out validation of novel sex-specific biomarkers as a key task to
fully understand reproductive cycle and influence of external factors (Anantharaman
and Craft, 2012; Sedik et al., 2010).
Reproductive cycle in mussels is annual and can be followed by the different
gonadal maturation stages (Hines et al., 2007). Male and female gonadal follicles
become empty after spawning up to the point of preventing sex identification by
histological or anatomical inspection. In Alfacs Bay, mussels were ripe between
October 2004 and January, with a main spawning period in January-February and a
second one of less intensity in April; although a new gametogenesis cycle began in
spring, it was interrupted by the high summer temperatures, and from July to September
gonads remained in a resting stage (Galimany et al., 2005). It is important to remark that
spawning takes places when temperatures are typically below 24°C since higher
temperatures increase M. galloprovincialis mortality, likely due to a limited capacity of
circulatory and ventilation systems that result in a reduction in the ability to assimilate
food and associated energy (Anestis et al., 2007). Therefore, coordination between
reproductive cycle and temperature is crucial to reproductive efficiency and adaptive
success of the species. Expression of VCLM7, one of the three major proteins from M.
edulis sperm (Takagi et al., 1994), matched very closely the maturation stage of male
mussels. High VCLM7 expression levels were observed in males during winter and
spring, just before and during the spawning events, whereas they decay in a period of
gonad degradation and inactivity (May and June). From May to December, the
progressive differentiation of the animals into low- and high VCLM7-producers
correspond to the progression of the gonadal development, from gonad inactivity to
gametogenesis and ripeness.
Since there is increasing evidence of bivalves impacted by endocrine disruptors
(Canesi et al., 2008; Ortiz-Zarragoitia and Cajaraville, 2006; 2008; Porte et al., 2006;
Oetken et al., 2004; Jobling et al., 2003), the development of new sex- and maturation
stage- biomarkers is pivotal to the use of mussels as monitoring species for
environmental risk assessment and ecotoxicology studies (Ciocan et al., 2012; Hines et
al., 2007; Saavedra and Bachere, 2006). VCLM7 follows male maturation process and
is pointed out as a potential suitable biomarker, supporting previous studies that present
VCLs as good candidates in this regard (Anantharaman and Craft, 2012).
4.3. Seasonal variations on stress genes biomarkers
4.3.1. Expression profiles of Fer and MT-10: variations in gradient of temperatureoxygen concentration
Fer and MT-10 expressions significantly correlated. In turn, increases in both
genes strongly correlated with increases in water temperature and decreases in dissolved
oxygen (Fig. 6). Since both environmental factors are intimately associated in aquatic
environments (Pörtner et al., 2007), it is difficult to discern the primary cause to explain
the observed effects. However, the fact that the correlation between GSTpi1 and both
Fer and MT-10 was only significant when temperature was not considered (Table S1,
Figures 5) suggests that oxidative stress may be the main cause behind the observed
expression patterns for both genes. The additional correlation with KRS may also point
in the same direction. The negative correlation of the expression of these genes with
salinity in the bays (after removing the seasonal/temperature factor) may indicate that
water conditions are a relevant factor on the physiology of the mussels in the Ebro bays.
Fer has been shown to play an important role in animal survival, including
mollusks, during periods with low oxygen concentrations (Larade and Storey, 2004). By
binding free iron, Fer prevents formation of highly toxic hydroxyl radicals both in phase
of hypoxia and aerobic recovery. During anoxia, exposure levels of ferritin heavy chain
transcripts are increased, although transcription regulation is produced in presence of
oxygen and not in anoxia conditions (Larade and Storey, 2004). In addition, Fer is upregulated by thermal stress which, in turn, may be also related with a temperaturedependent increase in ROS production as it was recently observed in Malaysian cockle
Tegillarca granosa (Jin et al., 2011) and red abalone Haliotis rufescens (Salinas-Clarot
et al., 2011). According to this, increases in ferritin content during warmer months
would be likely due to a secondary effect to gradual decrease in individuals aerobic
capacity.
MT-10 presents a very high basal expression and it is involved in essential metal
homeostasis, whilst MT-20 shows low basal expression and strong induction in
response to non-essential metals (Aceto et a., 2011; Fasulo et al., 2008; Vergani et al.,
2007; Dondero et al., 2005). Consequently, MTs are considered specific metalinduction biomarkers. However, metal levels in our sampling area were relatively low.
In agreement with its role, preliminary results disclosed hardly detectable levels of MT20 in our samples (data not shown), so it was not further considered in our study. No
correlation between MT-10 expression and metal content was observed neither, which
rule out exposure to metals as MT inductor in this case. Indeed, induction of MTs
expression has been correlated to other factors like hyperthermia (Gourgou et al., 2010)
and hyposmotic stress (Hamer et al., 2008). Similar to Fer, our results showed that MT10 was also sensitive to temperature and dissolved oxygen, which in turn may
contribute to assure the correct functioning of essential metals homeostasis, mainly Cu
and Zn, and prevent toxic free radicals production by leading protection and storage of
intracellular cations (Dondero et al., 2006). Temperature modulates MT-10 expression
in oysters through a heat shock element (HSE) located in the gene promoter, an element
also present in M. galloprovincialis promoter (Farcy et al., 2009; Piano et al., 2005;
Dondero and Viarengo, 2001). MTs may also show scavenging activity by direct
interaction with hydroxyl radicals (Viarengo et al., 1999); however, this protection
mechanism may be secondary, as some studies have shown no direct relationship
between oxidative stress and MT-10 (Dondero et al., 2005).
4.3.2. Oxidative stress gene biomarker: GSTpi1
Although GSTpi1, the most abundant GST isoform in mussels, has been
associated to detoxification of chemical compounds (Miao et al., 2011), field studies
questioned its role as chemical stress responsive gene (Hoarau et al., 2006). On the
contrary, GSTpi1 seems to be more responsive to oxidative stress in a process that
would be mainly mediated in terms of activity rather than transcription (Woo et al.
2013). In our study, GSTpi1 expression levels were variable depending on the season,
with strong negative correlation between GSTpi1 and temperature (Table 2), likely
evidencing the most prominent oxidative pressure in the coldest months (Di Salvatore et
al., 2013; Viarengo et al., 1991). However, the fact that GSTpi1 showed differences
depending on the differentiation stage, and strongly correlated with VCLM7 and actin,
the two genes related to the process of maturation, may indicate that reproduction cycle
is also important factor in GSTpi1 regulation (Schmidt et al., 2013).
GSTpi1 transcription levels also correlated with arsenic content in water (Table
2). Guidi et al., (2010) found that mollusks are able to bioaccumulate arsenic more
easily than other metals. As result, it may promote oxidative stress by reducing GSTpi1
activity and transcription (Chakraborty et al., 2013), which would explain the negative
correlation observed in our samples. On the contrary, Zn and Fe were shown as less
oxidative stress inductors, something already reported for other species (Devos et al.,
2012; Cantú-Medellín et al., 2009).
4.3.3. Genes related to metabolism of xenobiotics: CYP4Y1 and HSP70
CYP includes a family of genes, notably CYP1A, traditionally described to be
involved in phase I detoxification reactions of a variety of xenobiotic compounds in
many organisms, including aquatic species. However, although information is still very
limited, CYPs seem to play a different role in bivalves. CYP1-like genes have been very
recently identified and characterized in Mytilus (Zanette et al., 2013), but their
expressions were not sensitive to common AhR vertebrate agonists, the transcription
factor that regulates CYP1A expression. Also some studies have described CYP1A-like
activities and correlation with several chemicals, but presence of CYP1A gene could not
be certified in those cases (Chaty et al., 2004). As consequence, the attention has been
focused on other potential candidates responsible for the CYP1A-like activity, notably
CYP4Y1. CYP4Y1 is closely related to vertebrate CYP4A, CYP4F and CYP4T
subfamily members (Snyder, 1998), isoforms involved in fatty acid metabolism and
peroxisome proliferation, but also modulated by PAHs and PCBs (Reviewed in
Simpson, 1997). Indeed, mussel CYP4Y1 expression was inhibited after exposure to bnaphtoflavone (Snyder, 1998) and PAHs (Capello et al., 2013), and, therefore, shown to
be sensitive to xenobiotics. However, CYP4Y1 seems to be also involved in other
processes such as metabolism of fatty acid and prostaglandins. Also, Snyder et al.,
(2001) described reduction in CYP4Y1 mRNA in response to increasing percentage of
dissolved oxygen (DO), while acute hypoxia was not able to induce the opposite effect
(Woo et al., 2013). In our study, CYP4Y1 showed significant correlation with
temperature and other genes strongly correlated with temperature. Particularly
interesting was the negative correlation with HSP70, a gene believed to be induced by
temperature and DO, but not by PAHs (Izaguirre et al., 2014; Piano et al., 2004; Snyder
et al., 2001). Because concentration of oxygen in water is inversely related to
temperature, the positive and negative correlations of CYP4Y1 and HSP70,
respectively, may reflect the content of DO in the sampling area. Mussels increase the
duration of valve closure when increasing temperature, which may lead a shift from
aerobic to anaerobic metabolism (Anestis et al., 2008). In addition, expression levels of
CYP4Y1 and, more notably HSP70, were inversely correlated with river runoff values
(Table 2). This is consistent with the role of temperature/oxygen values on the
regulation of these genes, as low water inputs from the river associate to low oxygen
conditions in the areas of mussel culture.
5. Conclusion
Although mussels have characteristics that make them suitable bioindicators for
environmental risk assessment, there are still some drawbacks associated to their use
that may compromise the validity of some studies. One of the major concerns refers to
the lack of information on the mussel physiological status in specimens used for
biomonitoring studies. In this work, we have presented the tracing of several stressrelated gene expression biomarkers commonly used in biomonitoring (Fer, MT-10,
CYP4Y1, GSTpi1, HSP70 and KRS) during an annual life-cycle in cultured populations
of M. galloprovincialis with the aims of clarifying natural trends in the species and
prevent ambiguous or false conclusions in future studies. According to our results, all
biomarkers evaluated showed natural seasonal variations, in which temperature and
oxidative stress postulated as crucial modulators. As stated by Anestis et al. (2007;
2008), the high sensitivity to both environmental factors may reflect the proximity to
the acclimation limits in which mussels live, and explain the mortality events nonassociated to contamination eventually reported in the area. Interestingly, actin, a gene
commonly used as endogenous reference, was also shown as highly variable, so its use
as housekeeper is not recommended. Furthermore, actin expression seems to be closely
related to the maturation stage of the individual, as it is disclosed by comparison to the
male-specific gene VCLM7. As a whole, our results provide new bases for the
interpretation of the natural physiology in mussels and demonstrate that environmental
factors play important roles in the annual cycle of this group of organisms.
Conflict of interest
The authors confirm that there are no conflicts of interest.
Acknowledgements
We thank Mercedes Blázquez (ICM-CSIC, Barcelona) for her contribution to
facilitate original salinity and temperature data from the Ebro bays. This work has been
supported by Spanish Ministry of Economy and Competitiveness (CGL200801898/BOS and CTM2011-30471-C02-01) and the European Commission (Marine
Genomics Network of Excellence, GOCE-CT-2004-505403). S.J. has been also
partially supported by the EU Social Fund project in the Czech Republic No.
CZ.1.07/2.3.00/30.0009, OPVK program.
References
Aceto, S., Formisano, G., Carella, F., De Vico, G., Gaudio, L., 2011. The
metallothionein genes of Mytilus galloprovincialis: genomic organization, tissue
expression and evolution. Marine genomics 4, 61-68.
Albalat, A., Potrykus, J., Pempkowiak, J., Porte, C., 2002. Assessment of organotin
pollution along the Polish coast (Baltic Sea) by using mussels and fish as sentinel
organisms. Chemosphere 47, 165–71.
Anantharaman, S., Craft, J.A., 2012. Annual variation in the levels of transcripts of sexspecific genes in the mantle of the common mussel, Mytilus edulis. PLoS One 7,
e50861.
Anestis, A., Pörtner, H.O., Lazou, A., Michaelidis, B., 2008. Metabolic and molecular
stress responses of sublittoral bearded horse mussel Modiolus barbatus to warming
sea water: implications for vertical zonation. J. Exp. Biol. 211, 2889–98.
Anestis, A., Lazou, A., Portner, H.O., Michaelidis, B., 2007. Behavioral, metabolic, and
molecular stress responses of marine bivalve Mytilus galloprovincialis during
long-term acclimation at increasing ambient temperature. American journal of
physiology. Regulatory, integrative and comparative physiology 293, R911-921.
Banni, M., Negri, A., Mignone, F., Boussetta, H., Viarengo, A., Dondero, F., 2011.
Gene expression rhythms in the mussel Mytilus galloprovincialis (Lam.) across an
annual cycle. PLoS One 6, e18904.
Banni, M., Dondero, F., Jebali, J., Guerbej, H., Boussetta, H., Viarengo, A., 2007
Assessment of heavy metal contamination using real-time PCR analysis of mussel
metallothionein mt10 and mt20 expression: a validation along the Tunisian coast.
Biomarkers 12, 369–83.
Bebianno, M.J., Lopes, B., Guerra, L., Hoarau, P., Ferreira, A.M., 2007. Glutathione Stranferases and cytochrome P450 activities in Mytilus galloprovincialis from the
South coast of Portugal: effect of abiotic factors. Environ. Int. 33, 550–8.
Canesi, L., Borghi, C., Ciacci, C., Fabbri, R., Lorusso, L.C., Vergani, L., Marcomini,
A., Poiana, G., 2008. Short-term effects of environmentally relevant concentrations
of EDC mixtures on Mytilus galloprovincialis digestive gland. Aquat. Toxicol. 87,
272–9.
Cantú-Medellín, N., Olguín-Monroy, N.O., Méndez-Rodríguez, L.C., Zenteno-Savín,
T., 2009. Antioxidant enzymes and heavy metal levels in tissues of the black
chocolate clam Megapitaria squalida in Bahía de La Paz, Mexico. Arch. Environ.
Contam. Toxicol. 56, 60–6.
Cappello, T., Maisano, M., D’Agata, A., Natalotto, A., Mauceri, A., Fasulo, S., 2013.
Effects of environmental pollution in caged mussels (Mytilus galloprovincialis).
Mar. Environ. Res. 91, 52–60.
Chakraborty, S., Ray, M., Ray, S., 2013. Cell to organ: physiological, immunotoxic and
oxidative stress responses of Lamellidens marginalis to inorganic arsenite.
Ecotoxicol. Environ. Saf. 94, 153–63.
Chaty, S., Rodius, F., Vasseur, P., 2004. A comparative study of the expression of
CYP1A and CYP4 genes in aquatic invertebrate (freshwater mussel, Unio tumidus)
and vertebrate (rainbow trout, Oncorhynchus mykiss). Aquat. Toxicol. 69, 81–94.
Ciocan, C.M., Cubero-Leon, E., Peck, M.R., Langston, W.J., Pope, N., Minier, C.,
Rotchell, J.M., 2012. Intersex in Scrobicularia plana: transcriptomic analysis
reveals novel genes involved in endocrine disruption. Environ. Sci. Technol. 46,
12936–42.
Ciocan, C.M., Cubero-Leon, E., Minier, C., Rotchell, J.M., 2011. Identification of
reproduction-specific genes associated with maturation and estrogen exposure in a
marine bivalve Mytilus edulis. PLoS One 6, e22326.
Ciocan, C.M., Cubero-Leon, E., Puinean, A.M., Hill, E.M., Minier, C., Osada, M.,
Fenlon, K., Rotchell, J.M., 2010. Effects of estrogen exposure in mussels, Mytilus
edulis, at different stages of gametogenesis. Environ. Pollut. 158, 2977–84.
Cubero-Leon, E., Ciocan, C.M., Minier, C., Rotchell, J.M., 2011. Reference gene
selection for qPCR in mussel, Mytilus edulis, during gametogenesis and exogenous
estrogen exposure. Environ. Sci. Pollut. Res. Int. 19, 2728–33.
Devos, A., Voiseux, C., Caplat, C., Fievet, B., 2012. Effect of chronic exposure to zinc
in young spats of the Pacific oyster (Crassostrea gigas). Environ. Toxicol. Chem.
31, 2841–7.
Di Salvatore, P., Calcagno, J.A., Ortíz, N., Ríos de Molina, M.D.C., Sabatini, S.E.,
2013. Effect of seasonality on oxidative stress responses and metal accumulation in
soft tissues of Aulacomya atra, a mussel from the South Atlantic Patagonian coast.
Mar. Environ. Res. 92, 244–52.
Dondero, F., Dagnino, A., Jonsson, H., Capri, F., Gastaldi, L., Viarengo, A., 2006.
Assessing the occurrence of a stress syndrome in mussels (Mytilus edulis) using a
combined biomarker/gene expression approach. Aquatic Toxicology 78, S13-S24.
Dondero, F., Piacentini, L., Banni, M., Rebelo, M., Burlando, B., Viarengo, A., 2005.
Quantitative PCR analysis of two molluscan metallothionein genes unveils
differential expression and regulation. Gene 345, 259-270.
Dondero F, Viarengo A 2001. Identification of a putative metallothionein promoter
from the mediterranean mussel Mytilus galloprovincialis. 11 Int. Symp. on
pollutant responses in marine organisms, PRIMO 11, Plymouth UK, 10–13 July
2001. Abs 1102.
Farcy, E., Voiseux, C., Lebel, J.M., Fievet, B., 2009. Transcriptional expression levels
of cell stress marker genes in the Pacific oyster Crassostrea gigas exposed to acute
thermal stress. Cell stress & chaperones 14, 371-380.
Fasulo, S., Mauceri, A., Giannetto, A., Maisano, M., Bianchi, N., Parrino, V., 2008.
Expression of metallothionein mRNAs by in situ hybridization in the gills of
Mytilus galloprovincialis, from natural polluted environments. Aquat. Toxicol. 88,
62–8
Frank, S.N., Godehardt, S., Nachev, M., Trubiroha, A., Kloas, W., Sures, B., 2013.
Influence of the cestode Ligula intestinalis and the acanthocephalan Polymorphus
minutus on levels of heat shock proteins (HSP70) and metallothioneins in their fish
and crustacean intermediate hosts. Environ. Pollut. 180, 173–9.
Forbes, V.E., Palmqvist, A., Bach, L., 2006. The use and misuse of biomarkers in
ecotoxicology. Environ. Toxicol. Chem. 25, 272–80.
Garcia-Reyero, N., Raldua, D., Quiros, L., Llaveria, G., Cerda, J., Barcelo, D., Grimalt,
J.O., Pina, B., 2004. Use of vitellogenin mRNA as a biomarker for endocrine
disruption in feral and cultured fish. Analytical and Bioanalytical Chemistry 378,
670-675.
Gourgou, E., Aggeli, I.K., Beis, I., Gaitanaki, C., 2010. Hyperthermia-induced Hsp70
and MT20 transcriptional upregulation are mediated by p38-MAPK and JNKs in
Mytilus galloprovincialis (Lamarck); a pro-survival response. The Journal of
experimental biology 213, 347-357.
Guidi, P., Frenzilli, G., Benedetti, M., Bernardeschi, M., Falleni, A., Fattorini, D.,
Regoli, F., Scarcelli, V., Nigro, M., 2010. Antioxidant, genotoxic and lysosomal
biomarkers in the freshwater bivalve (Unio pictorum) transplanted in a metal
polluted river basin. Aquat. Toxicol. 100, 75–83.
Gracey, A.Y., Chaney, M.L., Boomhower, J.P., Tyburczy, W.R., Connor, K., Somero,
G.N., 2008. Rhythms of gene expression in a fluctuating intertidal environment.
Curr. Biol. 18, 1501–7.
Hamer, B., Jaksic, Z., Pavicic-Hamer, D., Peric, L., Medakovic, D., Ivankovic, D.,
Pavicic, J., Zilberberg, C., Schroder, H.C., Muller, W.E., Smodlaka, N., Batel, R.,
2008. Effect of hypoosmotic stress by low salinity acclimation of Mediterranean
mussels Mytilus galloprovincialis on biological parameters used for pollution
assessment. Aquatic toxicology 89, 137-151.
Hines, A., Yeung, W.H., Craft, J., Brown, M., Kennedy, J., Bignell, J., Stentiford, G.D.,
Viant, M.R., 2007. Comparison of histological, genetic, metabolomics, and lipidbased methods for sex determination in marine mussels. Anal Biochem 369, 175186.
Hoarau, P., Damiens, G., Roméo, M., Gnassia-Barelli, M., Bebianno, M.J., 2006.
Cloning and expression of a GST-pi gene in Mytilus galloprovincialis. Attempt to
use the GST-pi transcript as a biomarker of pollution. Comp. Biochem. Physiol. C.
Toxicol. Pharmacol. 143, 196–203.
Izagirre, U., Errasti, A., Bilbao, E., Múgica, M., Marigómez, I., 2014. Combined effects
of thermal stress and Cd on lysosomal biomarkers and transcription of genes
encoding lysosomal enzymes and HSP70 in mussels, Mytilus galloprovincialis.
Aquat. Toxicol. 149, 145–156.
Jin, C., Li, C., Su, X., Li, T., 2011. Identification and characterization of a Tegillarca
granosa ferritin regulated by iron ion exposure and thermal stress. Developmental
and comparative immunology 35, 745-751.
Jobling, S., Casey, D., Rodgers-Gray, T., Oehlmann, J., Schulte-Oehlmann, U.,
Pawlowski, S., Baunbeck, T., Turner, A.P., Tyler, C.R., 2003. Comparative
responses of molluscs and fish to environmental estrogens and an estrogenic
effluent. Aquatic Toxicology 65, 205-220.
Larade, K., Storey, K.B., 2004. Accumulation and translation of ferritin heavy chain
transcripts following anoxia exposure in a marine invertebrate. The Journal of
experimental biology 207, 1353-1360.
Luedeking, A., Koehler, A., 2004. Regulation of expression of multixenobiotic
resistance (MXR) genes by environmental factors in the blue mussel Mytilus
edulis. Aquat. Toxicol. 69, 1–10.
Manduzio, H., Cosette, P., Gricourt, L., Jouenne, T., Lenz, C., Andersen, O.K.,
Leboulenger, F., Rocher, B., 2005. Proteome modifications of blue mussel
(Mytilus edulis L.) gills as an effect of water pollution. Proteomics 5, 4958-4963.
Marigómez, I., Zorita, I., Izagirre, U., Ortiz-Zarragoitia, M., Navarro, P., Etxebarria, N.,
Orbea, A., Soto, M., Cajaraville, M.P., 2013. Combined use of native and caged
mussels to assess biological effects of pollution through the integrative biomarker
approach. Aquat. Toxicol. 136-137, 32–48.
Oetken, M., Bachmann, J., Schulte-Oehlmann, U., Oehlmann, J., 2004. Evidence for
endocrine disruption in invertebrates, International Review of Cytology, 1-44.
Orbea, A., Ortiz-Zarragoitia, M., Solé, M., Porte, C., Cajaraville, M.P., 2002.
Antioxidant enzymes and peroxisome proliferation in relation to contaminant body
burdens of PAHs and PCBs in bivalve molluscs, crabs and fish from the Urdaibai
and Plentzia estuaries (Bay of Biscay). Aquat. Toxicol. 58, 75–98.
Ortiz-Zarragoitia, M., Cajaraville, M.P., 2006. Biomarkers of exposure and
reproduction-related effects in mussels exposed to endocrine disrupters. Arch
Environ Con Tox 50, 361-369.
Ortiz-Zarragoitia, M., Cajaraville, M.P., 2008. Endocrine disruption in mussel
populations from the Biosphere's Reserve of Urdaibai (North Iberian Peninsula),
25th Congress of the European-Society-of-Comparative-Biochemistry-andPhysiology, Ravenna, Italy, S4-S4.
Pfaffl, M., 2001. A new mathematical model for relative quantification in real-time RTPCR. Nucl. Acids Res. 29, e45.
Pfaffl, M.W., Tichopad, A., Prgomet, C., Neuvians, T.P., 2004. Determination of stable
housekeeping genes, differentially regulated target genes and sample integrity:
BestKeeper--Excel-based tool using pair-wise correlations. Biotechnol Lett 26,
509-515.
Piano, A., Franzellitti, S., Tinti, F., Fabbri, E., 2005. Sequencing and expression pattern
of inducible heat shock gene products in the European flat oyster, Ostrea edulis.
Gene 361, 119-126.
Piano, A., Valbonesi, P., Fabbri, E., 2004. Expression of cytoprotective proteins, heat
shock protein 70 and metallothioneins, in tissues of Ostrea edulis exposed to heat
and heavy metals. Cell Stress Chaperones 9, 134–42.
Piña, B., Casado, M., Quirós, L., 2007. Analysis of gene expression as a new tool in
ecotoxicology and environmental monitoring. Trac-Trends in Analytical Chemistry
26, 1145-1154.
Porte, C., Janer, G., Lorusso, L.C., Ortiz-Zarragoitia, M., Cajaraville, M.P., Fossi, M.C.,
Canesi, L., 2006. Endocrine disruptors in marine organisms: Approaches and
perspectives. Comp Biochem Physiol C 143, 303-315.
Pörtner, H.O., Knust, R., 2007. Climate change affects marine fishes through the
oxygen limitation of thermal tolerance. Science 315, 95–7.
Puerto, M., Campos, A., Prieto, A., Cameán, A., de Almeida, A.M., Coelho, A.V.,
Vasconcelos, V., 2011. Differential protein expression in two bivalve species;
Mytilus galloprovincialis and Corbicula fluminea; exposed to Cylindrospermopsis
raciborskii cells. Aquat. Toxicol. 101, 109–16.
Puinean, A.-M., Rotchell, J.M., 2006. Vitellogenin gene expression as a biomarker of
endocrine disruption in the invertebrate, Mytilus edulis, Marine Environmental
Research.
Riginos, C., Wang, D., Abrams, A.J., 2006. Geographic variation and positive selection
on M7 lysin, an acrosomal sperm protein in mussels (Mytilus spp.). Mol Biol Evol
23, 1952-1965.
Saavedra, C., Bachere, E., 2006. Bivalve genomics. Aquaculture 256, 1-14.
Salinas-Clarot, K., Gutierrez, A.P., Nunez-Acuna, G., Gallardo-Escarate, C., 2011.
Molecular characterization and gene expression of ferritin in red abalone (Haliotis
rufescens). Fish & shellfish immunology 30, 430-433.
Schlesinger, M.J., 1990. Heat shock proteins. J. Biol. Chem. 265, 12111–4.
Schmidt, W., Power, E., Quinn, B., 2013. Seasonal variations of biomarker responses in
the marine blue mussel (Mytilus spp.). Mar. Pollut. Bull. 74, 50–5.
Sedik, W.F., Dempsey, K.E., Meng, X., Craft, J.A., 2009. Temporal expression of sexspecific genes in the mantle of the common mussel (Mytilus edulis). Mar. Biol.
157, 639–646.
Serafim, A., Lopes, B., Company, R., Cravo, A., Gomes, T., Sousa, V., Bebianno, M.J.,
2011. A multi-biomarker approac`h in cross-transplanted mussels Mytilus
galloprovincialis. Ecotoxicology 20, 1959–74.
Simpson, A.E., 1997. The cytochrome P450 4 (CYP4) family. Gen. Pharmacol. 28,
351–9.
Solé, M., Porte, C., Barcelo, D., Albaiges, J., 2000. Bivalves Residue Analysis for the
Assessment of Coastal Pollution in the Ebro Delta (NW Mediterranean). Mar.
Pollut. Bull. 40, 746–753.
Solé, M., Porte, C., Pastor, D., Albaigés, J., 1994. Long-term trends of polychlorinated
biphenyls and organochlorinated pesticides in mussels from the Western
Mediterranean coast. Chemosphere 28, 897–903.
Snyder, M.J., Girvetz, E., Mulder, E.P., 2001. Induction of marine mollusc stress
proteins by chemical or physical stress. Arch. Environ. Contam. Toxicol. 41, 22–9.
Snyder, M.J., 1998. Cytochrome P450 enzymes belonging to the CYP4 family from
marine invertebrates. Biochem. Biophys. Res. Commun. 249, 187–90.
Stuckas, H., Messerschmidt, K., Putzler, S., Baumann, O., Schenk, J.A., Tiedemann, R.,
Micheel, B., 2009. Detection and Characterization of Gamete-Specific Molecules
in Mytilus edulis Using Selective Antibody Production. Mol. Reprod. Dev. 76, 410.
Sureda, A., Box, A., Tejada, S., Blanco, A., Caixach, J., Deudero, S., 2011.
Biochemical responses of Mytilus galloprovincialis as biomarkers of acute
environmental pollution caused by the Don Pedro oil spill (Eivissa Island, Spain).
Aquat. Toxicol. 101, 540–9.
Takagi, T., Nakamura, A., Deguchi, R., Kyozuka, K.I., 1994. Isolation, characterization,
and primary structure of three major proteins obtained from Mytilus edulis sperm.
Journal of Biochemistry 116, 598-605.
Tanguy, A., Bierne, N., Saavedra, C., Pina, B., Bachere, E., Kube, M., Bazin, E.,
Bonhomme, F., Boudry, P., Boulo, V., Boutet, I., Cancela, L., Dossat, C., Favrel,
P., Huvet, A., Jarque, S., Jollivet, D., Klages, S., Lapegue, S., Leite, R., Moal, J.,
Moraga, D., Reinhardt, R., Samain, J.F., Zouros, E., Canario, A., 2008. Increasing
genomic information in bivalves through new EST collections in four species:
Development of new genetic markers for environmental studies and genome
evolution. Gene 408, 27-36.
Theil, E.C., 1987. Ferritin: structure, gene regulation, and cellular function in animals,
plants, and microorganisms. Annual review of biochemistry 56, 289-315.
Thellin, O., Zorzi, W., Lakaye, B., De Borman, B., Coumans, B., Hennen, G., Grisar,
T., Igout, A., Heinen, E., 1999. Housekeeping genes as internal standards: use and
limits. J Biotechnol 75, 291-295.
Venier, P., De Pitta, C., Pallavicini, A., Marsano, F., Varotto, L., Romualdi, C.,
Dondero, F., Viarengo, A., Lanfranchi, G., 2006. Development of mussel mRNA
profiling: Can gene expression trends reveal coastal water pollution? Mutat Res
602, 121-134.
Venier, P., Pallavicini, A., De Nardi, B., Lanfranchi, G., 2003. Towards a catalogue of
genes transcribed in multiple tissues of Mytilus galloprovincialis. Gene 314, 29-40.
Vergani, L., Grattarola, M., Grasselli, E., Dondero, F., Viarengo, A., 2007. Molecular
characterization and function analysis of MT-10 and MT-20 metallothionein
isoforms from Mytilus galloprovincialis. Arch. Biochem. Biophys. 465, 247–53.
Viarengo, A., Burlando, B., Cavaletto, M., Marchi, B., Ponzano, E., Blasco, J., 1999.
Role of metallothionein against oxidative stress in the mussel Mytilus
galloprovincialis. The American journal of physiology 277, R1612-1619.
Viarengo, A., Canesi, L., Pertica, M., Livingstone, D.R., 1991. Seasonal variations in
the antioxidant defence systems and lipid peroxidation of the digestive gland of
mussels. Comp. Biochem. Physiol. C. 100, 187–90.
Woo, S., Denis, V., Won, H., Shin, K., Lee, G., Lee, T.-K., Yum, S., 2013. Expressions
of oxidative stress-related genes and antioxidant enzyme activities in Mytilus
galloprovincialis (Bivalvia, Mollusca) exposed to hypoxia. Zool. Stud. 52, 15.
Woo, S., Jeon, H.-Y., Kim, S.-R., Yum, S., 2011. Differentially displayed genes with
oxygen depletion stress and transcriptional responses in the marine mussel, Mytilus
galloprovincialis. Comp. Biochem. Physiol. Part D Genomics Proteomics 6, 348–
356.
Zanette, J., Jenny, M.J., Goldstone, J. V, Parente, T., Woodin, B.R., Bainy, A.C.D.,
Stegeman, J.J., 2013. Identification and expression of multiple CYP1-like and
CYP3-like genes in the bivalve mollusk Mytilus edulis. Aquat. Toxicol. 128-129,
101–12.
Llebot, C., Sole, J., Delgado, M., Fernandez-Tejedor, M., Camp, J., Estrada, M., 2011.
Hydrographical forcing and phytoplankton variability in two semi-enclosed
estuarine bays. Journal of Marine Systems 86, 69-86.
Galimany, E., Ramón, M y Durfort, M (2005). Desarrollo gonadal del mejillón
Mytilus galloprovincialis de la bahía de Alfacs (delta del Ebro). X Congreso Nacional
de Acuicultura, Abstract Book II: 616-617.
Figure legends
Figure 1. (A) Correlation between expression of actin and water temperature. mRNA
levels are shown as normalized values. Black triangles, white circles and grey squares
indicate males, females and immature mussels, respectively. Regression line and the
corresponding correlation coefficient for all samples are included. (B) Effects of the
maturation stage in the expression of actin. The data are means of normalized values for
the mRNA expression levels of males and females pre-spawning (Pre-S) and postspawning (Post-S), and immature (I). Lowercase letters indicate homologous groups of
data in the Tukey's test (p < 0.05).
Figure 2. Changes in expression levels of VCLM7 (A) and actin (B) in M.
galloprovincialis collected during a complete yearly cycle. Ordinates indicate the month
of collection (1-January, 12-December). Empty circles, solid triangles and gray squares
indicate female, male and immature animals, respectively. Expression values are given
as copies of mRNA per 1000 copies of ribosomal S3 protein gene. Splines follow the
median expression ratio for males (discontinuous line) and females (continuous) (A),
and the median expression for all samples (discontinuous line) (B).
Figure 3. Analysis of PCA results. (A) Loading plots for PC1, PC2 and PC3; the
explained variation for each PC is indicated. (B) Score plot for PC1 and PC2. Males and
females are indicated by triangles and circles, respectively (black: pre-spawning; white:
post-spawning). Immature mussels are indicated by grey squares. Gene biomarkers and
temperature are shown as variables.
Figure 4. Effects of the maturation stage in the expression of MT-10, Fer, CYP4Y1,
GSTpi1, HSP70 and KRS. The data are means of normalized values for the mRNA
expression levels of males and females pre-spawning (Pre-S) and post-spawning (PostS), and immature (I). Lowercase letters indicate homologous groups of data (Tukey's
test, p < 0.05). N/S shows no significant differences between groups.
Figure 5. Correlograms between gene expression data and physical and chemical
parameters the Ebro delta bays. Top right, Pearson correlations; bottom left, partial
correlations after adjusting for temperature. Blue and red sectors indicate positive and
negative correlations, respectively. Note that the colored sector runs clockwise for
positive correlatios and anti- clockwise for negative ones. The extend of the colored
sectors indicate the strength of the correlation (See actual figures in Table S1)
Figure 6. Correlation between Fer and MT-10. Data are shown as normalized values.
Temperatures are indicated as °C in ranges: 5.00 - 9.99 (blue circles), 10 - 14.99 (green
diamonds), 15.00 - 19.99 (orange squares) and 20.00 - 24.99 (red triangles). Regression
line and the corresponding correlation coefficient for all samples are included. Blue and
red arrows indicate gradients for dissolved oxygen and temperature, respectively.