Usefulness of vitellogenin as a potential biomarker for endocrine disruption in the marine mussel mytilus galloprovincialis revisited from direct analysis at transcriptomic and proteomic level

Resumen

Since the mid-twentieth century, it has been observed that exposure to certain pollutants can cause alterations in organisms related to reproduction and behavior in many animals. These effects are the result of exposure to a group of chemicals called endocrine disruptors or EDCs (Colborn et al. 1996). EDCs involve a wide variety of compounds, including exogenous hormones and synthetic compounds that have different modes of action (Marty et al. 2011; Combarnous and Nguyen 2019). There are many types of endocrine disruption, remaining as one of the most studied ones the endocrine disruption caused by estrogenic EDCs (Sumpter 2005). This phenomenon has been widely studied in aquatic ecosystems after the discovery of intersex development in fish exposed to effluents from wastewater treatment plants (Arukwe et al. 1998; Desbrow et al. 1998), which are the main source of contribution of EDCs to aquatic ecosystems. In marine ecosystems, mussels of the genus Mytilus are one of the most used species in order to monitor marine pollution (Beyer et al. 2017). This is due to the characteristics that they possess: since they are sessile, filter feeders and are also widely distributed. In addition, they are capable of accumulating pollutants in their tissues without being severely affected, so they can be used as bioindicators of water pollution. For these reasons, mussels are one of the most used organisms for the monitoring of marine pollution, even developing monitoring programs based solely on the use of mussels, which are called “Mussel Watch Programs” (OSPAR Commission 2018). In the monitoring of marine pollution using mussels, two main aspects are combined: the analysis of the concentrations of pollutants in various matrices and the analysis of the different biomarkers. This allows detecting alterations in biological responses related to exposure to pollutants or their toxic effects (Vethaak et al. 2017). Currently, there are a great variety of biomarkers that are used in marine mussels; however, no specific biomarker has yet been developed to detect the effects of EDCs in these organisms. In order to try to overcome this matter, many authors have proposed vitellogenin (Vtg) levels as a biomarker of exposure to estrogenic EDCs. Vtg is the precursor of yolk proteins, which is normally produced in response to endogenous estrogens in mature female oviparous vertebrates, such as fish. Nevertheless, in general, in the case of males or juveniles, Vtg is present at very low levels compared to females (Wheeler et al. 2005; Hutchinson et al. 2006). Estrogenic EDCs have the ability to emulate estrogens so that they can bind to their receptors and trigger abnormal Vtg synthesis in male or juvenile fish. For this reason, in aquatic ecosystems, the induction of Vtg synthesis in male or juvenile fish has been used as a biomarker of exposure to estrogenic EDCs. The usefulness of this biomarker has been demonstrated in fishes both in field studies and in laboratory studies, in which the existence of a dose-response relationship between the levels of Vtg and the concentrations of estrogenic EDCs has been observed, as well as a positive relationship between the Vtg levels and exposure time (e.g. Funkenstein et al. 2000; Rose et al. 2002; Madsen et al. 2002; Zhang et al. 2004). At the present day, there are numerous direct and indirect methods for the measurement of the gene expression of Vtg, both at the transcript level and at the protein level, which have been used in fish to evaluate the effects of exposure to estrogenic EDCs (Hiramatsu et al. 2006; Hara et al. 2016). Among all these methods, the most widely used are immunological methods, such as immunological enzyme-linked immunosorbent assay (ELISA), which are based on obtaining specific antibodies for Vtg of the target specie. Due to the difficulty in obtaining antibodies for any species, other indirect methods have been proposed to estimate the amount of Vtg, such as the method of alkali labile phosphate or ALP that measures the phosphorous present in phosphorylated proteins. The success of this method to estimate Vtg levels lies in the fact that Vtg is a highly phosphorylated protein and is the most abundant in fish blood serum (Nagler et al. 1987; Yao and Crim 1996; Hallgren et al. 2009). Due to its great utility in estimating Vtg levels in fish, the ALP method was also adapted to be used in mollusks, especially in bivalves (Blaise et al. 1999). However, despite the fact that the synthesis of vitellogenin is a widely documented process in fish, in the case of mollusks, the biological basis of this process is still unknown. In spite of this, it has been assumed, from limited evidence, that the endocrine system of bivalve mollusks is similar to that of vertebrates such as fish; therefore, the Vtg synthesis process in mollusks would be homologous to that in vertebrates. For these reasons, Vtg has been proposed as a potential biomarker of estrogenicity (Matozzo et al. 2008) and ALP as a technique for its estimation (Blaise et al. 1999). However, unlike in fish, the results of ALP in bivalve mollusks are not conclusive, since the induction of Vtg was highly variable depending on the nature of the study, sex, exposure conditions or maturation stage of the organisms (e.g. Blaise et al. 1999; Ortiz-Zarragoitia and Cajaraville 2006; Aarab et al. 2006; Ricciardi et al. 2008; Baussant et al. 2011). This high variability in the obtained results has generated diverse opinions on the usefulness of Vtg as a biomarker of estrogenicity in mollusks, questioning its usefulness (Scott 2012, 2013). Therefore, this doctoral thesis has as its main objective to study the suitability of Vtg as a biomarker of endocrine disruption in the marine mussel M. galloprovincialis. For that purpose, the gene expression of Vtg is studied both at the transcript level and at the protein level after the exposure to a model estrogenic compound: the synthetic hormone 17α-ethinylestradiol (EE2). It is also determined whether there are differences between the gene expression patterns of Vtg at the transcript level and at the protein level by comparing both results. Additionally, taking into account that Vtg is a protein related to reproduction, it is also studied how they can affect both the nutritional status and the maturation status of individuals in the measurement of Vtg expression in M. galloprovincialis gonads. In order to achieve these objectives, several studies are carried out. The results of each of them are collected in chapters 2, 3 and 4 of this doctoral thesis. In Chapter 2, it is assessed whether the indirect technique of ALP, commonly used to estimate Vtg levels in bivalve mollusks, is a reliable technique. The use of this technique, by analogy with fish Vtg, is based on the assumption that Vtg is the most abundant phosphorylated protein in the tissue/fluid analyzed. A prior study to this thesis project showed that the use of ALP as a proxy for Vtg levels in the hemolymph of the freshwater mussel Elliptio complanata might not be adequate. In this case, the data obtained from ALP were compared with the direct measurement of Vtg, by ELISA (Morthorst et al. 2014). To carry out this study in which the gene expression of Vtg is evaluated at the protein level, a shotgun proteomic analysis was carried out for the identification and direct quantification of Vtg levels in the mature and immature gonad of females and males of M. galloprovincialis. Subsequently, the results obtained were compared with the Vtg levels estimated by the ALP method. Proteomic analysis showed that Vtg was only detected in female gonads and that Vtg expression levels vary depending on the state of the gonadal maturation of the sample. However, in the case of indirect estimation of Vtg levels using ALP, similar amounts of phosphorylated proteins were detected regardless the sex or the maturity stage of the gonads. Therefore, the results obtained in this study advise against the use of ALP as a proxy for Vtg levels in M. galloprovincialis gonads and, by extension, surely also in the rest of bivalve mollusks. In Chapter 3, once a reliable technique for the detection and quantification of Vtg in M. galloprovincialis gonads has been established, the possibility of Vtg being a suitable biomarker of estrogenicity in marine mussels, is explored. Prior studies to this thesis project suggested that Vtg was a suitable biomarker of estrogenicity in bivalve mollusks. The problem was that these studies used the ALP technique to estimate the Vtg levels of their experiments and, for that reason, their results may not be reliable. To carry out this study, three independent experimental exposures were carried out using individuals of the species M. galloprovincialis to the synthetic hormone EE2. In January 2015, a first exposure of 4 days duration to 10 and 100 ng / L EE2 was carried out, during which the mussels were not fed. In January 2016, a second exposure of 31 days duration to 100 ng/L EE2 was carried out, in which samples were taken at 24 -days-time and during which the mussels were fed a low diet. In November 2017, a third and final exposure was carried out for 24 days at 100 ng/L EE2, in which samples were taken at 4 and 24 days-time and which consisted of two experiments with two different regimens: one low (equal to the 2016’s experiment) and another high, in order to represent two different energy states. The proteomic analysis, as in Chapter 2, did not detect Vtg expression in male gonads, regardless EE2 concentration, regimen, or time of exposure. In the case of female gonads, Vtg expression levels were not affected after exposure to EE2, regardless the concentration, regime or time of exposure. However, in the case of females, a correlation between the degree of maturation of the gonad and the level of expression of Vtg was detected. The results of this study question the use of Vtg as a biomarker of endocrine disruption in M. galloprovincialis, and by extension in the rest of bivalve mollusks. Furthermore, it shows that the degree of maturation of the gonad can be an important confounding factor in the measurement of Vtg in the gonad of female mussels. Therefore, a histological study of the gonads is necessary before measuring the expression of Vtg. Otherwise, the natural biological variations in Vtg levels, due to the maturation stage of the gonad, could be interpreted with possible effects of estrogenic compounds. In Chapter 4, the possibility that Vtg is a suitable biomarker of estrogenicity in M. galloprovincialis gonads is also explored but, in this case, instead of evaluating the gene expression of Vtg at the protein level as it was done in the previous chapter, it is evaluated at the transcript level. This is due to the fact that some studies carried out in bivalve mollusks, such as the performed by Ciocan et al. (2010), showed an induction in the expression of Vtg at the transcript level after exposure to estrogenic compounds. Additionally, the fact that there is no induction at the protein level does not imply that there is no induction at the transcript level; this is due to various post-transcriptional regulatory events. Through this regulation, some transcripts may never be translated into their corresponding proteins, or even do so at different rates, so this could result in a lack of correlation between the abundance of some gene products at the transcript level and at the protein level (Vogel and Marcotte 2012; Diz and Calvete 2016). To carry out this study, the Vtg gene expression was analyzed at the transcript level using quantitative PCR (RT-qPCR) in samples from the high-regimen experiment performed in 2017. The analysis was performed on samples obtained from the same individuals in which the expression of Vtg at the protein level was analyzed in the previous chapter, in order to evaluate whether or not there is a correlation between the expression at the transcript level and at the protein level for Vtg in the gonad of M. galloprovincialis after exposure to EE2. Vitellogenin is a multi-domain protein and it has been described in so many occasions in both vertebrates and invertebrates that it can present different isoforms made up of different combinations of these domains. In the case of mollusks, there is still no solid evidence on the presence of isoforms for Vtg. For this reason, in this study the expression of the three conserved domains present in the Vtg of M. galloprovincialis was analyzed. In this way, the possible expression changes due to the formation of different isoforms of Vtg can be taken into account. In order to obtain reliable results in this Vtg gene expression study, a study of reference gene stability expression was carried out, and through it, two sets of independent genes were selected to perform the normalization of the expression data obtained for Vtg. Once the data were normalized, the results showed that in this case, Vtg gene expression is detected at the transcript level in both males and females, although the amount detected in females is much higher. The data also show that Vtg induction at the transcript level does not occur in any of the three domains after exposure to EE2, regardless of the sex and time of exposure. This, added to the results of the previous chapter, reinforces the hypothesis that Vtg is not an adequate biomarker of exposure to estrogens in M. galloprovincialis gonads and, by extension, it is probably not suitable in other bivalve mollusks either. Furthermore, the differences observed in the expression of the three analyzed domains suggest the possible presence of different Vtg isoforms in M. galloprovincialis. This chapter also makes a preliminary assembly of the complete Vtg sequences of Mytilus edulis and Mytilus trossulus, which until now, were not available. These sequences were assembled from transcripts obtained from transcriptome shotgun assembly (TSA) databases (Ramesh et al. 2019; Knöbel et al. 2020). This will allow the execution of similar studies to the one described in this chapter for these two species of the genus Mytilus. Finally, Chapter 5 presents the general discussion of the results of this doctoral thesis, as well as its implications. In first place, an overview of the use of ALP as a proxy for Vtg in bivalve mollusks is provided, ranging from its first use in 1999 to 2020. This review shows that despite the fact that there was already a prior study to this doctoral thesis that advised against the use of ALP as a proxy for Vtg in bivalves (Morthorst et al. 2014), no other study has questioned the use of ALP in these organisms and its use has not been discontinued over the years. A literature review is also conducted to evaluate the use of Vtg as an estrogenicity biomarker in marine bivalve mollusks. In this review, only studies carried out in bivalve mollusks that have used direct techniques to measure Vtg are taken into account. In this case, the lack of control of confounding factors in the measurement of Vtg is evidenced, such as sex maturity stage and the use of protandric organisms that can change their sex during the experiments if they have the adequate conditions. In this review, the absence of studies reporting negative results of xenoestrogenicity in bivalve mollusks is highlighted, as it is a fact that was previously discussed by another author (Scott 2013). In addition, the possible usefulness of other biomarkers of exposure to xenoestrogens, such as alterations in the expression of estrogen receptors, follicular atresia or alterations in the sex ratio, is also evaluated. Although, it seems that its usefulness as the one of the Vtg, is also questionable. This leads to suggest that, perhaps, the use of mollusks as species to evaluate the effects of endocrine disruptors typical on vertebrates is not appropriate. Therefore, the certainty that there is evidence both in favor and against the existence of a possible system homologous to that of vertebrates for the synthesis of steroids such as estradiol or testosterone in bivalve mollusks, is exposed. Currently, it cannot be assured that bivalve mollusks have the ability to synthesize vertebrate steroids such as estradiol or testosterone. Despite this, almost all the enzymes that are part of the vertebrate steroid synthesis pathway have been identified in various species of mollusks (Blalock et al. 2018; Thitiphuree et al. 2019). Other possible ways of regulating reproduction in mollusks are also described, in this case, based on current knowledge about different invertebrate systems such as insects and crustaceans, which may be more appropriate than those based on the ways of regulating reproduction in vertebrates. Finally, other possible approaches are indicated for the use of mussels for monitoring contamination by endocrine disruptors that are not based on vertebrate systems. This doctoral thesis constitutes one of the few studies that shows that ALP is not an adequate technique for estimating Vtg levels in marine mussels. Besides, it also suggests that Vtg may not be a suitable biomarker of xenoestrogenicity in marine mussels. All of this indicates that a further study of the functioning of the invertebrate endocrine system is necessary before it is possible to establish biomarkers of endocrine disruption.