Untitled

Environmental signaling and evolutionary change: can exposure of pregnant mammals to environmental estrogens lead to epigenetically Carlos Guerrero-Bosagna,a,b,Ã Pablo Sabat,b,c and Luis Valladaresa aLaboratorio de Hormonas y Receptores, Instituto de Nutricio´n y Tecnologı´a de los Alimentos (INTA), Universidad de Chile,Santiago, ChilebLaboratorio de Ecofisiologı´a Animal, Departamento de Ciencias Ecolo´gicas, Facultad de Ciencias, Universidad de Chile,Santiago, ChilecCenter for Advanced Studies in Ecology & Biodiversity and Departamento de Ecologı´a, Facultad de Ciencias Biolo´gicas,Pontificia Universidad Cato´lica de Chile, Santiago, ChileÃAuthor for correspondence: (email: [email protected]) DNA methylation is one of the epigenetic and environmental stimuli could produce effects, (ii) particular hereditary mechanisms regulating genetic expression in environmental agents as such stimuli, and (iii) that a genomic mammalian cells. In this review, we propose how certain persistent change be consequently produced in a population.
natural agents, through their dietary consumption, could Depending on the persistence of the environmental stimuli induce changes in physiological aspects in mammalian and on whether the affected genes are imprinted genes, mothers, leading to alterations in DNA methylation patterns induced changes in DNA methylation patterns could become of the developing fetus and to the emergence of new persistent. Moreover, some fragments could be more phenotypes and evolutionary change. Nevertheless, we frequently methylated than others over several generations, hypothesize that this process would require (i) certain key leading to biased base change and evolutionary con- periods in the ontogeny of the organism where the We believe that a first approach toward evaluating this problem requires separating the phenomenon of the emer- An old question in evolutionary biology is ‘‘how does var- gence of an evolutionary novelty into two processes: (i) that iation originate?’’ No matter how old this question is, the responsible for the origin of a new character and (ii) that controversy remains regarding whether (i) variability in pop- maintaining such a character over generations (i.e., fixation).
ulations appears exclusively by random mutations, a position Such separation has been previously proposed by authors defended by neo-Darwinism, or (ii) the formation of novel such as Futuyma and Moreno (1988) and West-Eberhard characters can, in some way, be induced by external environ- mental forces. Current understanding of epigenetic modifica- As Darwin, most evolutionary biologists have concentrat- tion of DNA shows that such controversy still exists. In this ed almost exclusively on the second process, that is the form sense, Jaenisch and Bird (2003) suggested that future lines of in which an evolutionary novelty can be fixed, not inquiring investigation should place emphasis on the identification of into the problem of how evolutionary novelties originate.
the stimuli that can initiate evolutionary changes. They pro- Variation among individuals and correlated differences in fit- posed that it is possible for external factors, such as dietary ness became a central topic in Darwin’s theory (Endler 1986) compounds, to lead to the accumulation of epigenetic changes and thereafter, Neo-Darwinian theory interpreted changes in over the years within populations. Given the recent evidences allelic frequencies of populations instead of studying the or- on mechanisms of epigenesis, here we propose that under igin of new phenotypes (Nijhout et al. 1986).
certain conditions, such epigenetic changes could become In accordance with the separation between origin and persistent over generations and this could have evolutionary fixation of an evolutionary novelty, some authors state that evolution is always a two-step process, first involving & BLACKWELL PUBLISHING, INC.
developmentally mediated variation, and then selection, At present, it is widely known that DNA methylation is whose operation results in gene frequency changes (Wake one of the epigenetic and hereditary aspects that regulate ge- and Larson 1987; West-Eberhard 1998). In this sense, changes netic expression in mammalian cells (Khosla et al. 2001).
arising because of alterations in early developmental process- Furthermore, DNA methylation is capable of being modified es, which, furthermore, could, in some cases, be environmen- by the action of externally applied agents (Mac Phee 1998).
tally induced, can appear whether or not such changes could Not all, but particular compounds found in nature could act become fixed and prosper in a population. Hence, in our as such agents. Moreover, they could be capable of affecting opinion, the diversity and evolution of species should be ex- the evolution of organisms, inducing profound changes in plained not only by those selective processes imposed by the individuals and populations, perhaps with transgenerational environment but also by the action of the environment as an consequences. We hypothesize that, whereas certain condi- inductor of genotypic and phenotypic variation, which is the tions are required for this process to occur, it is a feasible phenomenon. The task is to identify the conditions constrain- Regarding the persistence of such epigenetic changes through generations, long ago, Weismann (1893) stated that Experimental evidences concerning alterations of methyl- external influences may produce hereditary variations when ation patterns, at least in mammals, are generally restricted they are capable of modifying the determinants of the germ to studies of the effects of synthetic compounds or dietary plasm. Nevertheless, this could be only one of the ways restrictions of food items containing the methyl group (see through which environmental factors induce transgenerational Laird and Jaenish 1996; Singal and Ginder 1999). Although epigenetic changes. We recognize two ways for this to occur: this is very important for understanding the mechanisms of one is by dramatically modifying DNA aspects in the germ line DNA methylation, from an evolutionary perspective, it is of with transgenerational consequences, that is by means of pro- greater relevance to find compounds that are naturally in ducing mutations or transgenerationally persistent epigenetic contact with organisms; for example, those available for di- modifications in the genome, and the other is through inducing etary consumption, which, in addition, could produce alter- ontogenetical variation at every generation, although not pro- ations in patterns of DNA methylation in organisms.
ducing inheritance through the germ line. From our perspec- In this article, focusing exclusively on the phenomenon of tive, inductive environmental forces can act to create, through how evolutionary novelties originate, we describe how in one or both of these forms, new conformation of organisms, mammals, certain natural agents could induce alterations which also implies new possibilities within its surrounding in particular mechanisms of regulation of gene expression in environment. Jablonka and Lamb (1995) have named the individuals, such as methylation patterns, and the further range of the possible responses of individuals to new environ- arising of new, specific phenotypes in subsequent generations, mental challenges as the ‘‘reaction range’’ of individuals.
leading to evolutionary change. Nevertheless, we hypothesize Based on his experiments in Drosophila, Waddington pro- that this process would require (i) certain key periods in the posed two new concepts related to the capacity of environ- ontogeny of the organism where the environmental stimuli mental influences to induce the appearance of new characters could produce effects, (ii) particular environmental agents in organisms and their maintenance over generations. First, in as such stimuli, moreover, acting persistently, and (iii) that the face of disturbing and external stressing influences, there a persistent genomic change be consequently produced in a are counteracting tendencies in development toward normal adult conditions (i.e., canalization; Waddington 1959). Sec- The first requirement emerges because not all compounds ond, whereas these counteracting tendencies exist, if a stress- are capable of producing an effect on mothers that will have ing stimulus is capable of developmentally modifying a strain consequences on the fetus; the second emerges from the fact of organisms, the derived population may evolve exhibiting that an organism is not equally sensitive to outer stimuli the modification even in the absence of the stress (Wadding- throughout ontogeny; and the third because transgenerational ton 1952). He termed this process ‘‘genetic assimilation.’’ persistency of characters is ensured when it reaches the gen- An important fact to notice is that, through these concepts, omic level. Each of the three requirements presented will be Waddington distinguished particular environmental stimuli more extensively treated later in the text.
capable of inducing epigenetic changes, which are the ‘‘stress-ing’’ ones. McClintock (1984) also stated that a particularkind of stimuli producing stress lead to a genome’s reaction to it, whose response may underlie formation of new species.
Furthermore, she stated that genome produces programmedresponses, although it is necessary to subject the genome Experimentation on the problem of how evolutionary nov- repeatedly to the same challenge in order to observe the na- elties arise and the consequences on the genetic system of exposition to an environmental stimulus have been the focus of epigenetic studies in a variety of organisms, includ- There are multiple isoforms of Dnmts, but all are encoded ing Drosophila (Rutherford and Lindquist 1998), bacteria by the same cytosine–Dnmt gene (Deng and Szyf 1998).
(Cairns et al. 1988), and yeast (Steele and Jinks-Robertson Among these isoforms, Dnmt1o is a variant of Dnmt1 that accumulates in oocyte nuclei during the follicular growth Several types of epigenetic inheritance have been described phase, and Dnmt3L is an isoform of Dnmt3a and Dnmt3b, to date. Jablonka and Lamb (1995) have proposed three sys- but that lacks Dnmt enzymatic activity and interacts with tems of epigenetic inheritance: (i) steady-state systems, such as Dnmt2a and Dnmt3b (Kierzenbaum 2002). Dnmt3L acts as a Wright’s (1945) persistence of alternative cellular states as a cofactor for de novo methylation of imprinted genes in the result of changes in nuclear genes or in cytoplasmic constit- female gametes and for the establishment of methylation im- uents of the cell, (ii) structural inheritance systems, such as the prints in oocytes (Hata et al. 2002).
maintenance through generations of the ciliary patterns in It is worth noting that Dnmt1 is localized principally in protozoa, albeit of the genetic constitution of the cells in- somatic cell nuclei, but it is cytoplasmatic in the oocyte and in volved (Nanney 1985), and (iii) chromatin-marking systems, or the preimplantation embryo (Bestor 2000). However, the var- those related to the transmission of specific patterns of the iant Dnmt1o has transient nuclear localization in the eight-cell chromatin structure (Holliday 1987; Jablonka et al. 1987).
stage, corresponding to the time when genomic imprints are Specifically, the latter refers to non-DNA parts of the chro- established (Howell et al. 2001). On the other hand, Dnmt3L mosomes that are capable of binding proteins or additional co-localizes with Dnmt3a and Dnmt3b in mammalian cell chemical groups attached to DNA bases, which affect the nature and stability of gene expression, now commonly Given the crucial role of the diverse Dnmts in the epige- named genomic imprinting. DNA methylation describes a netic modification of DNA, it is of great interest to know postreplicative modification, in which a methyl group is add- whether there are environmental substances capable of mod- ed to a DNA residue in a covalent manner (Laird and Jaenish ifying the intracellular levels of such enzymes or their patterns 1996); for this reason, it is a form of genomic imprinting. The of gene expression. Nevertheless, no studies have reported this DNA methylation reaction is enzymatically catalyzed by kind of interaction, which we suspect may have a role in DNA methyltransferases (Dnmts) and takes place in 50 to 30- relating environmental stimuli to DNA modification. How- oriented CG dinucleotides, which are known as CpG sites, at ever, the recent finding that individual Dnmts can be tracked, the carbon 5 of the cytosine ring (Singal and Ginder 1999).
and that their binding to genomic DNA can be quantified in CpG islands are regions with a high frequency of CpG sites; vivo in mammalian cells (Liu et al. 2003) can be enormously these islands are often associated with genes, and are usually helpful for determining the link between environmental com- found in promoter zones (Gardiner-Garden and Frommer pounds and the process of DNA methylation.
1987). CpG sites are not evenly distributed within the genome,and are preferentially unmethylated, regardless of the tran-scriptional activity of the associated gene (Bird 1986). As other regions are normally methylated, patterns of genomic DNA methylation can be distinguished along the genome (Singal and Ginder 1999; Bestor 2000; Jones and Takai 2001).
Nevertheless, there is controversial information regarding Roemer et al. (1997) were the first to show reappearance in whether methylation patterns are established because of the progeny of modified characters in parents. In their exper- the enzymatic activity of one or more Dnmts (Bestor 2000; iments on rodents, the adult phenotype produced because of the fusion of pronuclei with eggs of different genotypes was There are at least three families of Dnmts described to also observed in the offspring. Furthermore, such transgene- date: Dnmt1, Dnmt2, and Dnmt3. However, there is no rational persistence of the modified characters was related to agreement regarding whether each one plays a specific, dif- altered methylation patterns that were, in turn, transmitted ferential role in the process of DNA methylation (Bestor through male gametogenesis. However, not all genes are 2000). It has been speculated that Dnmt3A and Dnmt3B are equally capable of passing on changes in patterns of methyl- responsible for the establishment of methylation patterns ation. There is a particular class of genes, crucial for under- during early development, whereas Dnmt1 is responsible for standing the mechanisms of epigenetic inheritance, that are the further maintenance of such patterns. Experiments con- known to have relatively unchanged methylation patterns ducted in vitro support this model, revealing that Dnmt1 has over generations. These genes, named ‘‘imprinted genes’’, do a preference for hemimethylated DNA as a substrate (Yoder not seem to be affected by overall alterations in methylation et al. 1997), whereas Dnmt3A and Dnmt3B act as a de novo patterns that take place early in development (Constaˆncia et methyltranferase, preferring unmethylated DNA (Yokochi al. 1998). Such genes carry a molecular memory of their pa- rental origin that is acquired early in the germ line (Surani 2001). This molecular memory is associated with specific me- However, even if no one imprinted gene is affected when thylation patterns in CpG islands of each allele, which con- altered by an environmental signal, environmentally induced sequently affect further genic expression (Costello and Plass changes in methylation patterns could also become persistent if such changes, and the environmental conditions allowing Once the allelic differences in methylation of imprinted the establishment of such changes, are both conserved genes are defined (during the establishment of germinal line throughout generations. This could occur whenever there is in the developing embryo), such differences generally remain a concordance, an association between the environmental stable in the somatic tissues (Constaˆncia et al. 1998). The stimuli, the established DNA methylation patterns, and the marking process of these genes appears to involve three stag- resulting phenotype of an organism. For instance, if some es: (i) the establishment of marks in gametes; (ii) the perma- natural agent can induce the loss of methylation in genes and nence of these marks during embryogenesis and in the adult produce phenotypic alterations (e.g., those modifications somatic tissues; and (iii) the erasure of marks in the early germ emerging from the loss of methylation in Igf2), a standard line (Razin and Cedar 1994). Conclusive information on the phenotypic pattern will arise every time the specific environ- way in which methylation in imprinted genes is initiated from mental stimuli lead to the establishment of particular patterns an unmethylated state during gametogenesis is still elusive of methylation. Still, it is important to consider that this could (Ferguson-Smith and Surani 2001). However, recent investi- be a broader phenomenon, and environmentally induced gations indicate that primordial germ cells are substantially changes in methylation patterns could affect several other methylated (which corresponds to the same pattern in somatic imprinted genes as well. As a result, an environmental stim- cells) before they colonize gonads and become demethylated ulus would bias the phenotypic change toward certain types around the time of entry into the gonads (Hajkova 2002). An incomplete deletion of marks during gametogenesis would Nevertheless, the consequences of altering DNA methyl- explain the inheritance of the parental epigenotype (Reik et al.
ation toward specific persistent patterns could imply mutation in those specific segments of the genome. For instance, itis known that a methylated cytosine is half-way to thesubstitution of a cytosine for a thymidine. The completion ofconversion requires only a hydrolytic deamination reaction (Singal and Ginder 1999). Therefore, if some methylated sites are frequently methylated over several generations, it is pos-sible that an eventual base change from cytosine to thymidine Imprinted genes may be susceptible to undergoing changes in will occur more frequently than any other substitution. In methylation patterns during preimplantational development fact, CpG sites are hotspots for transitions from cytosine to (Khosla et al. 2001). As imprinted genes tend to conserve thymidine, generated by a spontaneous deamination of 5-me- methylation patterns from one generation to the next, chang- thyl cytosine to thymidine (Coulondre et al. 1978). The result ing methylation patterns in these genes could lead to the ap- would be, as mentioned by West-Eberhard (2003), that pearance of the derived alterations in the future generations.
‘‘evolved sensitivity to environmental influence during gene Therefore, if external agents are capable of inducing partic- expression could influence susceptibility to certain kinds of ular changes in methylation patterns in these genes, such structural change during evolution.’’ changes could flourish transgenerationally. Moreover, thiscould take place in the absence of the stimuli that initiallychanged its methylation pattern, generating a process that would be a kind of Waddington’s ‘‘genetic assimilation’’ but Changes of methylation patterns in certain imprinted genes The first condition for our statement on environmentally in- can generate associated specific phenotypes (see Morison et duced evolution is that the process must occur early in on- al. 2001 for examples). Particularly interesting, from our per- togeny, before or during the establishment of the germ line in spective, is the Beckwith–Wiedemann syndrome. Researchers metazoa. This is important for two main reasons: first, be- suspect that this syndrome is related to the loss of imprinting cause eventual reprogramming of methylation patterns in the in Igf2, and is characterized by somatic overgrowth, macro- germ line can be transmitted to the progeny (Surani 2001), glossia, abdominal wall defects, visceromegaly, and an in- and second, because during development, there is an en- creased susceptibility to childhood tumors (Caspary et al.
hanced susceptibility of the organism to the action of outer 1999). Therefore, in this case, a change in methylation pat- compounds, with greater consequences in the adult than when terns in a single gene can lead to phenotypic changes in several the same stimulus occurs later in ontogeny (Amzallag 2000).
With regard to the latter statement, Gould and Lewontin (1979) have emphasized that during the early ontogenetic that allow such organization to take place (Maturana- stages of complex organisms, ‘‘differentiation of organ sys- Romesı´n and Mpodozis 2000). An experimental approach tems and their integration into a functioning body is such a to such a statement comes from Clark and Galef (1995), who delicate process so easily derailed by early errors, with accu- proposed that daughters tend to resemble their mothers not only because both share a relatively large proportion of their The morphogenic process of an organism is basically the genes but also because they tend to have similar histories of product of a three-way interaction between the environment, genetic factors, and those characteristics that emerge from Applying this view to DNA methylation, reproduction a self-organized dimension created by development itself plays a key role in passing on those changes in patterns of (Amzallag 2000). The establishment of methylation patterns methylation that could eventually arise during early stages of during early development (as well as other processes in the ontogeny. Reproduction, in addition to conserving the morphogenesis) also depends on the immediate environment pattern of DNA methylation of an organism’s genome experienced by the embryo. These methylation patterns will throughout generations in a lineage, will also conserve the guide the formation of particular cell types by controlling conditions allowing such patterns of methylation to be es- gene expression (Holliday 1998), therefore biasing further tablished in every generation. Hence, for a mammal to be formed from a zygote, and for development to take place In mammals, patterns of methylation are established for generating a phenotype similar to the parental phenotypic the entire genome at least three times during development.
pattern, the process requires not only the genetic content that The periods in which reprogramming of methylation patterns provides a zygote with the potential to become an adult but takes place are: (i) before the implantation of the embryo, (ii) also a surrounding environment for the embryo, which en- during the development of the germ line (Reik et al. 2001), sures the occurrence of appropriate methylations, at key pe- and (iii) during the period beginning soon after blastocyst riods of time during the embryological process.
implantation (Constaˆncia et al. 1998) until gastrulation (Mac Nevertheless, in mammals, despite the fact that the uterus Phee 1998). Before blastocyst implantation, a great part of the acts as a buffer for either mechanical or chemical perturbat- DNA is demethylated (Dean et al. 1998); thus, the DNA of ions on the developing embryo, making the developmental blastocysts hardly shows methylation (Mac Phee 1998).
process more isolated from environmental perturbations than Between blastocyst implantation and gastrulation, there is a in other taxa, the development is still susceptible to particular wave of de novo methylations that restore the overall me- perturbations. Maternal effects such as variations in the thylation patterns, which is retained in the somatic cells of hormonal status of a mother are capable of affecting the animal for the rest of its life (Mac Phee 1998). In the germ microenvironment in which the fetus develops (Clark and line, reprogramming takes place by overall demethylations Galef 1998) and, consequently, its later ontogenetic processes and methylations of the genome (Constaˆncia et al. 1998). In (Bernardo 1996). For example, studies have shown that mice, primordial germ cells undergo an overall demethylation differential exposition to hormones can affect characters of process in early development until day 13 or 14 (Reik et al.
the embryo. Clark et al. (1993) and Vandenbergh and Hug- 2001). Later, during gametogenesis, there is a de novo me- gett (1994) demonstrated that the intrauterine position of thylation event until the previously observed high levels of female rodents affects the sex ratio of their litters, which is methylation in the zygote (Mac Phee 1998), oocyte, and because of differential prenatal exposure to steroidal hor- sperm genomes (Reik et al. 2001) are reached. It is likely that mones, which in turn depends on the gender of neighboring both demethylations taking place during the first stages of postzygotic cleavage, and methylations occurring after im- Besides, the hormonal state of the mammalian female can plantation, are important in removing acquired epigenetic be strongly influenced by the environment through com- modifications, especially those acquired during gametogenesis pounds that are naturally found in her diet (Nagao et al.
2001). In this sense, it has been reported that feed toxicants, ordietary imbalances of specific nutrients, can alter the compo-sition of oviductal and uterine secretions (McEvoy et al.
Thus, the establishment of methylation patterns in the embryo is a process that depends directly on the environmentin which it takes place, that is the intrauterine environment, Reproduction involves the conservation in the progeny not but also indirectly on the surrounding environmental signa- only of the structure required to carry out the self-conserved ling, which, in some way, alters such an intrauterine environ- organization represented by the organism but also the pres- ment. Accordingly, perturbing the intrauterine environment ervation of the structural characteristics of the environment while early development takes place could bring about consequences in the establishment of methylation patterns, could also interpret this cell transformation as alterations in with the corresponding phenotypic repercussions.
methylation patterns. Some evidence for this phenomenoncomes from studies in chicken liver, where estrogens appear toact in the regulation of expression of the vitellogenin I and II, and VLDL II genes, through changes in patterns of methyl- ation of estrogen-responsive element sites (Edinger et al.
1997). It has also been shown that neonatal exposure to DESand adult ovary hormones produces abnormalities in the de- Our second condition is that only particular compounds in methylation of the lactoferrine promoter, which shows that nature could act as environmental inputs for environmentally either hormonal xenobiotics or natural hormones are capable induced evolution to take place. The early embryo is exqui- of triggering impairments during the development of organs sitely sensitive to alterations in its environment (McEvoy et al.
2001). Nevertheless, not every compound with which a mam- It has been reported that environmental estrogens can also malian mother has contact in nature is capable of altering the produce direct effects on DNA methylation patterns. For ex- embryo environment, although some compounds could lead ample, administration of the phytoestrogens cumestrol and to alterations in mammalian hormonal features. Furthermore, equal to newborn mice can enhance methylation and produce we believe that some environmental compounds can, in ad- inactivation in the proto-oncogene H-ras (Lyn-Cook et al.
dition to altering the hormonal status of a mammalian moth- 1995). In addition, Day et al. (2002) demonstrated that me- er, be in turn capable of affecting important processes during thylation patterns can be altered in 8-week-old mice that the early development, including the establishment of me- consumed high quantities of genistein.
thylation patterns in the embryo. Among those environmen- With respect to hormonal effects early in development, tally available compounds capable of affecting the hormonal Holliday (1998) was the first to envisage a possible link be- status of a mammalian mother, there are some of synthetic tween hormone action and establishment of DNA methylat- origin, or xenobiotics (Danzo 1998) and of natural origin, ion in mammalian embryos. He proposed that the effect of such as phytoestrogens. The latter refers to secondary met- teratogens on a mother might disrupt the normal distribution abolites produced by plants (Croteau et al. 2000; Yu et al.
of DNA methylation in a developing fetus, producing devel- 2000) that produce estrogenic action at a variety of levels in opmental abnormalities or defects that can appear in the animals (McLachlan 2001). Phytoestrogens are readily avail- subsequent generations. Newbold et al. (2000) reported that able in the environment for animal consumption and their after administering DES to pregnant rats during early post- physiological, hormonal, and nonhormonal effects in animals implantational development and neonatality, a greater sus- have been studied to some extent (Levy et al. 1995; Santell ceptibility for specific tumor formation in rete testis and et al. 1997; Boettger-Tong et al. 1998; Milligan et al. 1998; reproductive tract tissues occurred in F1 and appeared further Gallo et al. 1999). Some phytoestrogens such as genistein and in the non-DES exposed F2. These authors speculated that daidzein belong to a class of flavonoids, the so-called isoflav- this transgenerational phenomenon could implicate epigenetic ones (Liggins et al. 2000). The consumption of isoflavones can alterations that were transmitted through germ line, including elicit uterotrophic and mammatrophic effects in mice and on changes in methylation patterns. Although this finding strong- the hypothalamic/pituitary axis as well (Santell et al. 1997). In ly suggests alteration and further transmission of a genomic humans, it has been reported that the consumption of phyt- change through germ line across more than one generation in oestrogens affects levels of the sex hormone-binding globulin, response to an early exposition to an estrogenic compound, which regulates the bioavailability of steroidal sex hormones there is still missing evidence on the mechanism behind this process and whether it implies changes in DNA methylation Changing the hormonal status in mammals could have consequences beyond the physiologic level. McLachlan (2001) In the experiments of Newbold et al. (2000), the trans- suggested that estrogens could play a role in programming or generational persistence of the enhanced susceptibility to imprinting those genes involved in cell proliferation, differen- tumor formation takes place when mothers are exposed to tiation, or survival, either directly or through related signaling DES after embryo implantation; however, estrogens play an pathways. He also proposed that an estrogenic chemical may important role even before implantation occurs. The implan- directly imprint a gene through a process leading to persistent tation process involves complex interactions between the genetic change, probably at the level of DNA methylation. In blastocyst and the uterus (Paria et al. 1993). Uterine preim- this sense, Barrett et al. (1981) suggested that diethylstilbestrol plantational estrogen secretions are essential for activating the (DES), a powerful estrogenic synthetic compound, could blastocyst of Mus musculus for further implantation, which transform cells by mechanisms other than punctual muta- is not possible if estrogen secretions are prevented by tions, frameshift mutations, or small deletions. Currently, one ovariectomization (Paria et al. 1998). Nevertheless, just as estrogenic stimuli are needed for normal development, pre- cysts, which could occur through membrane-mediated estro- implantational exposure to synthetic estrogenic compounds gen actions, directly induced by isoflavones in uterine can lead to phenotypic alterations. For instance, Takai et al.
secretions, or mediated by other compounds secreted in the (2000) reported that in utero preimplantational exposure of uterine epithelia such as 4-OH-17b-estradiol. The formation rodent embryos to the synthetic estrogen bisphenol-A leads to of this compound in uterine epithelia could be related to an increased body mass of the animals at weaning. Further- plasmatic isoflavone content, although no studies have at- more, Wu et al. (2004) have recently shown that in vitro early tempted to detect such compounds in uterine secretions, or exposure to the environmental contaminant 2,3,7,8-tetra- showed that its high consumption can alter the production of chlorodibenzo-p-dioxin can indeed alter DNA methylation cathecolestrogens in the uterine epithelia.
patterns in preimplantational embryos. Interestingly, those Although there is strong evidence suggesting that the hor- genes changed, H19 and IGF-2, were imprinted genes.
monal status of mammalian mothers can be an important Although it is not known whether compounds with est- feature related to the establishment of methylation patterns in rogenic action (CEA) inside the uterus could act directly upon early embryos, so far, there is no concluding evidence of this.
the developing embryo, or via intermediaries, it is possible We believe that an investigation on this subject should be that the relationship between estrogenic stimuli and methyl- performed in order to uncover the aspects behind an eventual ation in the preimplantational embryo is mediated by the ex- epigenetic role of estrogenic compounds (both animal pro- pression of c-fos. While on the one hand it is known that c-fos duced, plant produced, and synthetic) on developmental directly regulates the dnmt1 transcription, increasing Dnmt1 processes, in particular, on the establishment of methylation levels (Bakin and Curran 1999), on the other, the induction of c-fos is a response attributed to membrane-mediated estrogenactions (Das et al. 2000). Through this mechanism, whichprovides an alternative pathway to the classical estrogen re- THE ‘‘GENOMIC CHANGE’’ REQUIREMENT FOR A ceptors a and b, CEA could trigger responses, as has been observed in pancreatic b cells (Nadal et al. 2000). In summary,the membrane-mediated estrogenic actions would first induce The third requirement that we propose for the environmental c-fos and then trigger the activation of the Dnmt1 enzyme.
and hormonal induction to become an evolutionary process is Furthermore, in blastocysts, this indirect and membrane- that genomic change should be achieved. Evolutionary mediated relationship between estrogenic stimuli and c-fos change in the morphogenetic process must arise from chang- activation could also occur. In preimplantational blastocysts, es in patterns of regulation and interaction during ontogeny Paria et al. (1998) demonstrated that latent blastocysts can be (see discussion by Atchley 1987). Such a connection gains activated if they are incubated in vitro with 4-OH-17b-estra- special importance when considering that the patterns of reg- diol, a catecholestrogen synthesized from 17b-estradiol in ulation and interaction occurring at early stages in ontogeny uterine luminal epithelia by the action of the hydrogen-2/hid- could, even in mammals, be susceptible to environmental roxilase-4 enzyme. This response to 4-OH-17b-estradiol could also occur via a pathway distinct from the classical nuclear Nevertheless, the question arising at this point goes beyond estrogen receptors (Paria et al. 1998). In addition, Paria et al.
the relationship between the environmental stimuli and even- (1998) found that 4-OH-17b-estradiol increases with the ep- tual epigenetic consequences on DNA methylation. The chal- ithelial growth factor (EGF) receptor. Interestingly, other lenge is to know how an eventual change in DNA studies have demonstrated that an increase in the EGF re- methylation patterns could become persistent and evolution- ceptor may also be related to activation of c-fos (Kamiya et al.
ary. Besides, another question arises, regarding the definition 1996). On the other hand, a direct induction of c-fos by est- of evolutionary change. Is persistence in the conditions al- rogen has also been shown in different cell types (Allen et al.
lowing the establishment of changed methylation patterns 1997; Garcia et al. 2000), which occurs via an estrogen re- across lineages a sufficient attribute for such changes to be ceptor element present in this gene (Hyder et al. 1992). Thus, considered as evolutionary, or do such changes need to reach estrogenic stimuli could induce c-fos, either directly, through the threshold of mutation at the genomic level? a gene receptor, or indirectly through membrane-mediated It is true that genomic mutational change ensures a great degree of persistence through generations. However, persist- Furthermore, phytoestrogens could also induce c-fos and ence can also be the result of two processes, as previously consequently alter methylation patterns in cells. A study sup- mentioned: (i) the environment could persistently trigger, porting this view demonstrated that the intake of genistein in generation after generation, a specific change in methylation ovariectomized female rodents induced the expression of the patterns, or (ii) persistence could be present in intrinsic fea- RNA messenger of c-fos in the uterus (Santell et al. 1997).
tures of the organisms as, for instance, the stable nature of Hence, we suspect that phytoestrogens can also act on blasto- Given the special feature of imprinted genes regarding In this sense, if a natural population of rodents is suddenly possessing methylation patterns that are more stable across subjected to a high intake of phytoestrogens, it is feasible generations than other genes, persistence could be achieved to hypothesize that such a high intake by pregnant rodents through changing methylation patterns of imprinted genes. In could influence the normal reproductive process, altering this view, such changes in imprinted genes could have the same the mother’s hormonal status, the intrauterine signaling, evolutionary value of mutations, given that there is an asso- and, consequently, the establishment of DNA methylation ciated character variation with the changes, and because of the patterns in embryos. The resulting phenotypes will be in ac- persistence of these changes throughout generations. Thus, the cordance with the particular pattern of DNA methylation definition of ‘‘evolutionary change’’ at this point becomes achieved as a consequence of the environmental stimuli, blurred. What is true is that persistence through generations represented in this case by phytoestrogens. As a result, such could be achieved in alternative ways to genomic mutation.
changes in methylation patterns will persist in the population Nevertheless, speaking in terms of genomic mutation, this if the organisms are constantly subjected to this same could be achieved when the persistent change in methylated environmental input and consequently, the achieved pheno- cytosines bias to specific mutations, as previously mentioned.
types will also persist throughout generations. Nevertheless, it Regarding the frequency of eventual mutations derived is important to point out that such a newly formed phenotype from changes in methylation patterns, given that such must not be considered to be associated with any adaptive changes can be environmentally induced, they cannot be con- goal; on the contrary, the new forms of organisms will fit sidered to be at random. Therefore, we can expect that in these within the environment they live, resulting from an environ- cases, the appearance of mutation will be in greater frequency mental input that leads to standardized phenotypes in than when mutation is considered to be at random. In fact, concordance with the environmental stimuli that produced there is a 12-fold higher than normal mutation rate for the them. In the particular case of imprinted genes, changing conversion of the methylated form of CpG to TpG and CpA, methylation patterns on those genes could imply transgene- which reduces the occurrence of CpG to about 20% of its rational persistence of epigenetic changes in the absence of expected frequency in vertebrate genomes (Sved and Bird 1990).
the environmental input that initially produced them.
Despite the evidence suggesting that the environment, Because the early stages of ontogeny play key roles in the through the action of naturally consumed agents, can alter establishment of phenotypic variation, it is important to the developmental process to the point that the emerging determine how environmental signals (particularly CEA) are alterations can be inherited as evolutionary change, conclusive involved in the developmental process. Nevertheless, a com- information is still elusive. Evidence in the direction of plete understanding of this involvement is difficult at this genomic change derived from alterations in methylation time. One of the complications is that the mechanisms patterns is needed for our hypothesis to be plausible in the through which estrogen and CEA bring about physiological classic view of the meaning of evolutionary change.
actions are not yet clearly understood (Nilsson et al. 2001).
Further studies on the effects of CEA on organisms, especiallyduring early stages of ontogeny, are needed to provide new insights, and to help in the understanding of the impact of this class of compounds on ecosystems in general (McLachlan 2001), and, particularly, on the physiologically relevant evo-lutionary processes that guide the formation of organisms and Holliday (1998) proposed that teratogens could target mech- anisms that control patterns of DNA methylation in partic-ular regions of the genome of developing embryos, modifying methylation patterns of the same DNA sequence in somatic The authors wish to thank Roxanna Gomez, Paula Neill, and Renee cells, leading to a developmental alteration, and subsequently Hill for their help in the preparation of the manuscript and linguisticrevision. We also thank Dr. Gloria Calaf for her critical review of the producing changes in germ line cells. Moreover, if such manuscript and valuable suggestions. Carlos Guerrero-Bosagna altered methylation patterns are eventually transmitted to a gratefully acknowledges CONICYT and also INTA and Nestle´ subsequent generation, the same type of defect might be seen (Holliday 1998). Phytoestrogens could act in the same man-ner, but with the peculiarity that they are naturally available for consumption by many organisms. Phytoestrogens arepresent in high quantities in food items commonly included in Allen, D., Mitchner, N., Uveges, T., Nephew, K., Khan, S., and Jonathan, the natural dietary composition of rodents, such as fruits, N. 1997. Cell-specific induction of c-fos expression in the pituitary glandby estrogen. Endocrinology 138: 2128–2135.
nuts, seeds (Liggins et al. 2000) and especially wheat, oats, Amzallag, G. N. 2000. Connectance in sorghum development: beyond the genotype–phenotype duality. BioSystems 56: 1–11.
Atchley, W. 1987. Developmental quantitative genetics and the evolution of Gardiner-Garden, M., and Frommer, M. 1987. CpG islands in vertebrate ontogenies. Evolution 41: 316–330.
genomes. J. Mol. Biol. 196: 261–282.
Bakin, A., and Curran, T. 1999. Role of DNA methylcytosine transferase in Gould, S. J., and Lewontin, R. C. 1979. The spandrels of San Marco and cell transformation by fos. Science 283: 387–390.
the Panglossian paradigm: a critique of the adaptationist programme.
Barrett, C., Wong, A., and McLachlan, J. 1981. Diethylstilbestrol induces Proc. Roy. Soc. Lond. B205: 581–598.
neoplastic transformation without measurable gene mutation at two loci.
Hajkova, P., et al. 2002. Epigenetic reprogramming in mouse primordial germ cells. Mech. Dev. 117: 15–23.
Bernardo, J. 1996. Maternal effects in animal ecology. Am. Zool. 36: Hata, K., Okano, M., Lei, H., and Li, E. 2002. Dnmt3L cooperates with the Dnmt3 family of de novo DNA methyltransferases to establish maternal Bestor, T. 2000. The DNA methyltransferases of mammals. Hum. Mol.
imprints in mice. Development 129: 1983–1993.
Holliday, R. 1987. The inheritance of epigenetic defects. Science 238: Bird, A. P. 1986. CpG-rich islands and the function of DNA methylation.
Holliday, R. 1998. The possibility of epigenetic transmission of defects in- Boettger-Tong, H., et al. 1998. A case of a laboratory animal feed with high duced by teratogens. Mutat. Res. 422: 203–205.
estrogenic activity and its impact on in vivo responses to exogenously Howell, C. Y., et al. 2001. Genomic imprinting disrupted by a maternal administered estrogens. Environ. Health Perspect. 106: 369–373.
effect mutation in the Dnmt1 gene. Cell 104: 829–838.
Cairns, J., Overbaugh, J., and Miller, S. 1988. The origin of mutants. Nature Hyder, S., Stancel, G., Nawaz, Z., McDonnell, D., and Loose-Mitchell, D.
1992. Identification of an estrogen response element in the 30- flanking Caspary, T., Cleary, M. A., Perlman, E. J., Zhang, P., Elledge, S. J., and region of the murine c-fos protooncogene. J. Biol. Chem. 267: 18047– Tilghman, S. M. 1999. Oppositely imprinted genes p57(Kip2) and igf2 interact in a mouse model for Beckwith–Wiedemann syndrome. Genes Jablonka, E., Goitein, R., Sperling, K., Cedar, H., and Marcus, M. 1987. 5- aza-C-induced changes in the time of replication of the X chromosomes Clark, M., and Galef, B. 1995. Prenatal influences on reproductive life- of Microtus agrestis are followed by non-random reversion to a late history strategies. Trends Ecol. Evol. 10: 151–153.
pattern of replication. Chromosoma 95: 81–88.
Clark, M., and Galef, B. 1998. Perinatal influences on the reproduc- Jablonka, E., and Lamb, M. 1995. Epigenetic Inheritance and Evolution, the tive behavior of adult rodents. In T. A. Mousseau and C. W. Fox (eds.).
Lamarkian Dimension. Oxford University Press Inc., New York.
Maternal Effects as Adaptations. Oxford University Press, New York, Jaenisch, R., and Bird, A. 2003. Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals. Nat.
Clark, M., Karpluk, P., and Galef, B. 1993. Hormonally mediated inher- itance of acquired characteristics in Mongolian gerbils. Nature 364: Jones, A., and Takai, D. 2001. The role of DNA methylation in mammalian epigenesis. Science 293: 1068–1070.
Constaˆncia, M., Pickard, B., Kesley, G., and Reik, W. 1998. Imprinting Kamiya, K., Sato, T., Nishimura, N., Goto, Y., Kano, K., and Iguchi, T.
mechanism. Genome Res. 8: 881–900.
1996. Expression of estrogen receptor and proto-oncogene messenger Costello, J., and Plass, C. 2001. Methylation matters. J. Med. Gen. 38: ribonucleic acids in reproductive tracts of neonatally diethylstilbestrol- exposed female mice with or without post-pubertal estrogen administra- Coulondre, C., and Miller, J. 1978. Molecular basis of base substitution tion. Exp. Clin. Endocrinol. Diabetes 104: 111–122.
hotspots in Escherichia coli. Nature 274: 775–780.
Khosla, S., Dean, W., Brown, D., Reik, W., and Feil, R. 2001. Culture of Croteau, R., Kutchan, T., and Lewis, N. 2000. Natural products (secondary preimplantation mouse embryos affects development and the expression metabolites). In B. Buchanan, W. Gruissem, and R. Jones (eds.). Bio- of imprinted genes. Biol. Reprod. 64: 918–926.
chemistry and Molecular Biology of the Plants. Courier Companies Inc., Kierzenbaum, A. 2002. Genomic imprinting and epigenetic reprogramming: unearthing the garden of forking paths. Mol. Reprod. Dev. 63: 269–272.
Danzo, B. 1998. The effects of environmental hormones on reproduction.
Laird, P., and Jaenisch, R. 1996. The role of DNA methylation in cancer Cell. Mol. Life Sci. 54: 1249–1264.
genetics and epigenetics. Annu. Rev. Genet. 30: 441–464.
Das, S., Tan, J., Raja, S., Halder, J., Paria, B., and Dey, S. 2000. Estrogen Levy, J., Faber, K., Ayyash, L., and Hughes, C. 1995. The effect of prenatal targets genes involved in protein processing, calcium homeostasis, and exposure to the phytoestrogen genistein on sexual differentiation in rats.
wnt signaling in the mouse uterus, independent of estrogen receptor-a Proc. Soc. Exp. Biol. Med. 208: 60–66.
and -b. J. Biol. Chem. 275: 28834–28842.
Li, S., et al. 1997. Developmental exposure to diethylstilbestrol elicits de- Day, J., et al. 2002. Genistein alters methylation patterns in mice. J. Nutr.
methylation of estrogen-responsive lactoferrin gene in mouse uterus.
Dean, W., et al. 1998. Altered imprinted gene methylation and expression in Liggins, J., Bluck, J., Runswick, S., Atkinson, C., Coward, W., and Bing- completely ES cell-derived mouse fetuses: association with aberrant phe- ham, S. 2000. Daidzein and genistein content of fruits and nuts. J. Nutr.
notypes. Development 125: 2273–2282.
Deng, J., and Szyf, M. 1998. Multiple isoforms of DNA methyltransferase Liu, K., Wang, Y. F., Cantemir, C., and Muller, M. 2003. Endogenous are encoded by the vertebrate cytosine DNA methyltransferase gene.
assays of DNA methyltransferases: evidence for differential activities of DNMT1, DNMT2, and DNMT3 in mammalian cells in vivo. Mol. Cell.
Edinger, R., Mambo, E., and Evans, M. 1997. Estrogen-dependent tran- scriptional activation and vitellogenin gene memory. Mol. Endocrinol. 11: Lyn-Cook, B. D., Blann, E., Payne, P. W., Bo, J., Sheehan, D., and Med- lock, K. 1995. Methylation profile and amplification of proto-oncogenes Endler, J. 1986. Natural Selection in the Wild. Princeton University Press, in rat pancreas induced with phytoestrogens. Proc. Soc. Exp. Biol. Med.
Ferguson-Smith, A., and Surani, M. A. 2001. Imprinting and the epigenetic Mac Phee, D. 1998. Epigenetics and epimutagens: some new perspectives asymmetry between paternal genomes. Science 293: 1086–1089.
on cancer, germ line effects and endocrine disrupters. Mutat. Res. 400: Futuyma, D., and Moreno, G. 1988. The evolution of ecological special- ization. Annu. Rev. Ecol. Systems 19: 207–233.
Maturana-Romesı´n, H., and Mpodozis, J. 2000. The origin of species by Gallo, D., et al. 1999. Reproductive effects of dietary soy in female Wistar means of natural drift. Rev. Chil. Hist. Nat. 73: 261–300.
rats. Food. Chem. Toxicol. 37: 493–502.
McClintock, B. 1984. The significance of responses of the genome to chal- Garcia, E., Lacasa, D., and Giudicelli, Y. 2000. Estradiol stimulation of c-fos and c-jun expressions and activator protein-1 deoxyribonucleic McEvoy, T. G., Robinson, J. J., Ashworth, C. J., Rooke, J. A., and Sinclair, acid binding activity in rat white adipocyte. Endocrinology 141: K. D. 2001. Feed and forage toxicants affecting embryo survival and fetal development. Theriogenology 55: 113–129.
McLachlan, J. 2001. Environmental signalling: what embryos and evolu- Santell, R. C., Chang, Y. C., Muralee, G. N., and Helferich, W. G. 1997.
tion teach us about endocrine disrupting chemicals. Endocr. Rev. 22: Dietary genistein exerts estrogenic effects upon the uterus, mammary gland and the hypothalamic/pituitary axis in rats. J. Nutr. 127: Morison, I. M., Croydon, J. P., and Cleverly, S. D. 2001. The imprinted gene and pattern-of-origin effect database. Nucleic Acids Res. 29: Singal, R., and Ginder, G. 1999. DNA methylation. Blood 93: 4059–4070.
Steele, F., and Jinks-Robertson, S. 1992. An examination of adaptive re- Milligan, S. R., Balasubramanian, A. V., and Kalita, J. C. 1998. Relative version in Saccharomyces cerevisiae. Genetics 132: 9–21.
potency of xenobiotic estrogens in an acute in vivo mammalian assay.
Surani, M. A. 2001. Reprogramming of genome function through epige- Environ. Health Perspect. 106: 23–26.
netic inheritance. Nature 414: 122–128.
Nadal, A., Ropero, A., Laribi, O., Maillet, M., Fuentes, E., and Soria, B.
Sved, J., and Bird, A. 1990. The expected equilibrium of the CpG dinuc- 2000. Nongenomic actions of estrogens and xenostrogens by binding at a leotide in vertebrate genomes under a mutation model. Proc. Natl. Acad.
plasma membrane receptor unrelated to estrogen receptor a and estrogen receptor b. Proc. Natl. Acad. Sci. USA 97: 11603–11608.
Takai, Y., et al. 2000. Preimplantation exposure to bisphenol A advances Nagao, T., Yoshimura, S., Saito, Y., Nakagomi, M., Usumi, K., and Ono, postnatal development. Reprod. Toxicol. 15: 71–74.
H. 2001. Reproductive effects in male and female rats of neonatal ex- Thigpen, J., et al. 1999. Phytoestrogen content of purified, open- and closed- posure to genistein. Reprod. Toxicol. 15: 399–411.
formula laboratory animal diets. Lab. Anim. Sci. 49: 530–536.
Nanney, D. L. 1985. Heredity without genes: ciliate explorations of clonal Vandenbergh, J., and Huggett, C. 1994. Mother’s prior intrauterine position heredity. Trends Genet. 1: 295–298.
affects the sex ratio of her offspring in house mice. Proc. Natl. Acad. Sci.
Newbold, R., Hanson, R. B., Jefferson, W. N., Bullock, B. C., Haseman, J., and McLachlan, J. A. 2000. Proliferative lesions and reproductive tract Waddington, C. H. 1952. Genetic assimilation of an acquired character.
tumors in male descendants of mice exposed developmentally to diet- hylstilbestrol. Carcinogenesis 21: 1355–1363.
Waddington, C. H. 1959. Canalization of development and genetic assim- Nijhout, F. H., Wray, G. A., Kremen, C., and Teragawa, K. 1986. On- ilation of acquired characters. Nature 183: 1654–1655.
togeny, phylogeny and evolution or form: an algorithmic approach.
Wake, D., and Larson, A. 1987. Multicellular analysis of evolving lineage.
Nilsson, S., et al. 2001. Mechanisms of estrogen action. Physiol. Rev. 81: West-Eberhard, M. J. 1998. Evolution in the light of developmental and cell biology, and vice versa. Proc. Natl. Acad. Sci. USA 95: 8417–8419.
Paria, B., Huet-Hudson, Y., and Dey, S. 1993. Blastocyst’s state of activity West-Eberhard, M. J. 2003. Developmental Plasticity and Evolution. Oxford determines the ‘‘window’’ of implantation in the receptive mouse uterus.
Proc. Natl. Acad. Sci. USA 90: 10159–10162.
Weismann, A. 1893. The Germ-Plasm: A Theory of Heredity. AMS Press Paria, B., Lim, H., Wang, X., Liehr, J., Das, S., and Dey, S. 1998. Co- ordination of differential effects of primary estrogen and catecholestro- Wright, S. 1945. Genes as physiological agents. General considerations. Am.
gen on two distinct targets mediates embryo implantation in the mouse.
Wu, Q., Ohsako, S., Ishimura, R., Suzuki, J. S., and Tohyama, C. 2004.
Pino, A., Valladares, L., Palma, M., Mancilla, A., Yan˜ez, M., and Albala, C. 2000. Dietary isoflavones affect sex hormone-binding globulin levels chlorodibenzo-p-dioxin (TCDD) alters the methylation status of im- in postmenopausal women. J. Clin. Endocr. Metab. 85: 2797–2800.
printed genes H19 and Igf2. Biol. Reprod. 70: 1790–1797.
Razin, A., and Cedar, H. 1994. DNA methylation and genomic imprinting.
Yoder, J. A., Soman, N. S., Verdine, G. L., and Bestor, T. H. 1997. DNA (cytosine-5)-methyltransferases in mouse cells, and tissues. Studies with a Reik, W., Dean, W., and Walter, J. 2001. Epigenetic reprogramming in mechanism-based probe. J. Mol. Biol. 270: 385–395.
mammalian development. Science 293: 1089–1092.
Yokochi, T., and Robertson, K. D. 2002. Preferential methylation of un- Roemer, I., Reik, W., Dean, W., and Klose, J. 1997. Epigenetic inheritance methylated DNA by mammalian de novo DNA methyltransferase in the mouse. Curr. Biol. 7: 277–280.
Dnmt3a. J. Biol. Chem. 277: 11735–11745.
Rutherford, S. L., and Lindquist, S. 1998. Hsp90 as a capacitor for mor- Yu, O., et al. 2000. Production of the isoflavones genistein and daidzein in phological evolution. Nature 396: 336–342.
non-legume dicot and monocot tissues. Plant Physiol. 124: 781–793.

Source: http://www.pablosabat.cl/wp-content/uploads/2013/02/2005-Environmental-signaling-and-evolutionary-change-can-exposure-of.pdf