Epigenetics can be defined as a change in phenotype that is heritable but does not involve a change in the nucleotide sequence in DNA; that is, a change in genotype. This definition is very broad encompassing a variety of phenomena.
Epigenetic changes during cellular differentiation
For example, a change in phenotype of a single cell that is then passed on to its descendants qualifies as an epigenetic phenomenon. Thus it includes the various pathways of differentiation that are taken by cells during the embryonic development of an organism. Examples:
- X-inactivation — where one of the two X chromosomes in female mammals is inactivated in each cell early in development and that same chromosome remains inactivated in all the descendants of that cell.
- Imprinting — where whether a gene in a cell lineage is expressed or not depends on which parent contributed the gene.
The great stumbling block in converting differentiated cells into induced pluripotent stem cells (iPSCs) was to find ways of reversing the epigenetic changes in the differentiated cell (e.g., a skin cell) to unlock its full developmental potential. Stable changes in gene expression are brought about in two main ways:
- DNA methylation — where its cytosines are methylated. This usually represses the activity of that DNA.
- Histone modifications — where methyl, acetyl, and other groups are added to the histones in chromatin. Prominent examples:
- adding methyl groups to the #4 lysine in histone H3 ("H3K4me"). This is associated with active genes in that region of the chromatin.
- adding methyl groups to the #27 lysine in histone H3 ("HeK27me"). This is associated with gene silencing.
- epigenetic "writers": enzymes that add chemical groups to histones or DNA.
- epigenetic "erasers": enzymes that remove these groups.
- epigenetic "readers": proteins that recognize specific epigenetic modifications of histones or DNA producing a change in gene expression, e.g., increasing (or decreasing) gene transcription.
Epigenetic changes induced in her embryos by the mother
Pregnant mice fed a diet enriched in chemicals that donate methyl groups (e.g., folic acid, vitamin B12) alter the expression of a coat color gene (agouti) in their offspring even though no change in the gene structure occurs. Presumably gene expression is suppressed by the increased methylation of CpG dinucleotides at the gene locus. However, note that this phenomenon does not involve a change in gene expression that is passed through the germline from parent to offspring. In fact, we would not expect this to occur because imprints in mammals are usually erased between generations by the removal of methyl groups from the DNA in the cells that will form the gametes.
Some epigenetic changes pass from parents to offspring.
In Linaria vulgaris
Linaria vulgaris (Figure 9.5.1) is a member of the snapdragon family and goes by the common name of toadflax or butter-and-eggs. Back in 1749, the father of taxonomy, Linnaeus, noted that while most members had typical bilaterally-symmetrical snapdragonlike flowers, occasionally plants were found that produced radially-symmetrical flowers instead. Today we know that its flower symmetry is under the control of a gene, Lcyc. We also know that the nucleotide sequence of Lcyc is identical in both the bilateral and radial flower types. What is different is the pattern of methylation of this gene. Lcyc in the normal flowers is only lightly methylated while the gene in the plants producing radially-symmetrical flowers is heavily methylated. The heavily methylated gene gets passed on to subsequent generations, although occasionally it reverts to the less-methylated form and then the plants produce normal flowers.
Figure 9.5.1: Linaria vulgaris in a meadow. Image used with permission (CC BY 2.0; ArcziPL).
Several intriguing patterns of inheritance in Drosophila do appear to result from epigenetic changes that are passed along in sperm and/or eggs; that is, in the germline. When fed the antibiotic geldanamycin, a mutant strain of Drosophila forms bristly outgrowths on its eyes. When mated, this trait appears in the subsequent generations (as many as 13 when the experiment was stopped) even though they are never again exposed to geldanamycin. There is no evidence that any mutation had occurred. There is very little methylated DNA in the Drosophila genome so this phenomenon probably requires histone modifications.
Vinclozolin is a commercial fungicide that when ingested by mammals is converted into compounds that bind to androgen receptors in the nucleus. When pregnant rats are injected with vinclozolin, their male offspring show reduced fertility: they produce fewer sperm and these swim poorly. When these males do succeed in mating, their male offspring continue to show the same defects and the problem passes along in the male germline at least for two more generations. There is no evidence that any mutation has occurred, but there does appear to have been an alteration in the pattern of DNA methylation.
Young female mice reared in an enriched environment (in a big cage with regularly-changed plastic play tubes, cardboard boxes, running wheel, various pet toys, and nesting material) develop enhanced long-term potentiation (LTP). When returned to a normal unenriched cage environment and mated, not only does their enhanced LTP persist, but their offspring, growing up in a conventional unenriched laboratory cage, display enhanced LTP as well.
When female inbred mice raised on a normal diet are mated with male mice of the same inbred strain but raised on a low-protein diet, the expression of hundreds of genes in their offspring differs from that in offspring of mice where both parents were raised on a normal diet. The males were kept with the females only for mating so the only contribution that they could make to their offspring must have been carried in their sperm.
The Inheritance of Acquired Characteristics?
Back in 1809, the French naturalist Jean-Baptiste Lamarck postulated that traits acquired during the lifetime of an organism could be passed on to its offspring. Thus he explained the long neck of giraffes as resulting from years of straining to reach foliage on the trees overhead. With the discoveries of genetics, his theory was discarded. It didn't depend solely on that. After all, male Jewish babies have been circumcised for millennia, but there has been no reduction in the size of the foreskin in all that time. So to paraphrase the words of Hamlet, "There is a destiny that shapes our ends rough-hew them how we will".
We have seen examples above of the phenotypic expression of genes being altered by the environment and this alteration continues to be seen in subsequent generations even in the absence of the environmental influence that created it. For epigenetic changes induced in a female, the change must persist in her great grandchildren (F3) to qualify as inheritance of an acquired characteristic. This is because a pregnant female (F0) carries not only her F1 fetus but the fetus itself (F2) has already set aside its own germ cells. So an environmentally-induced epigenetic change to DNA could have affected all three generations. For males (F0), appearance of the change in grandchildren (F2) will qualify because only himself and his sperm (F1) have been exposed to the stimulus. So the passing of an environmentally-induced epigenetic change from one generation to the next may or may not qualify as the inheritance of an acquired characteristic.
What about humans?
A few studies have produced suggestive evidence that environmental influences (e.g., smoking, diet) on grandparents can lead to phenotypic changes in their grandchildren. These were retrospective studies with their built-in limitations. Correlation does not guarantee causation. If the initial exposure was to a female, we must see the trait in her greatgrandchildren to qualify.
In the human fetus (and in mice as well), a massive wave of DNA demethylation occurs in the cells — called primordial germ cells — that will later form sperm and eggs. So how to reconcile this erasure of epigenetic marks with evidence of their persisting from one generation to the next? Perhaps the answer is to be found in the small number of genes that escape demethylation, and these could be the basis of the rare cases of epigenetic inheritance across generations.