Category Archives: primers

Genomic Imprinting VI: Hemimethylation

So, last time, we discussed the fact that the expression differences associated with genomic imprinting rely on the existence of epigenetic differences, such as DNA methylation. We also mentioned that those differences are established separately in the male and female germ lines. That is, one methylation pattern is established in the female germ line during oogenesis (egg formation), while a different pattern is established in the male germ line during spermatogenesis (sperm formation).

It is straightforward to understand how such differences could be established, since oogenesis and spermatogenesis occur in physically distinct locations, where different patterns of gene expression can produce the epigenetic differences. But, after fertilization, these parent-of-origin-specific epigenetic marks are maintained across many rounds of cell division. So, a cell in, say, your liver, will exhibit different epigenetic states on maternally and paternally alleles, despite the fact that they have occupied the same cellular environment throughout development.

Alleles at an imprinted locus maintain substantial epigenetic differences, despite occupying the same environment across many cell divisions.

The allele-specific maintenance of the methylation state depends on the fact that methylation occurs at “palindromic” sequences. When we’re talking about language, a palindrome is a word or phrase that contains the same sequence of letters when read forwards or backwards, like “a man, a plan, a canal, Panama,” or “eat tea.” In genetics, a palindrome is where the nucleotide sequence on one strand of the DNA is the same as the sequence on the complementary strand (which is read in the opposite direction).

The palindrome we will be concerned with here is a really short one: CpG (where the “p” indicates the phosphate linker between the cytosine (C) and guanine (G) nucelotides).  CpG is a palindrome because C pairs with G (and G pairs with C), so that the complementary DNA strand has a CpG at the same site. Methylation occurs on the cytosines, so that if we have two alleles with different methylation states at a CpG site, they will look like this:

Here we have a schematic representation of two alleles with palindromic CpG sites that differ in their epigenetic state. In the top figure, the cytosines on both strands of the double-stranded DNA are methylated. In the bottom figure, both cytosines are unmodified.

So, how is this methylation difference maintained when the cell undergoes DNA replication and mitosis? The key lies in the fact that DNA replication is semi-conservative. That is, in order to make a copy of a double-stranded piece of DNA, what you do is pull the two strands apart and synthesize a new strand complementary to each of them.

When Watson and Crick published their paper on the structure of the DNA double helix, they noted that this structure suggested a mechanism by which the specific DNA sequence could be replicated. Crick later went on to establish a reputation for himself as a neuroscientist. Watson went on to establish a reputation for for himself as an asshole.

The newly-synthesized strands will contain normal, unmethylated cytosine, whether or not the template strand was methylated.  So, starting from unmethylated DNA, each daughter cell inherits an unmethylated copy of the allele. But, if we start from methylated DNA, each daughter cell inherits hemimethylated DNA, where one strand of the DNA double helix has a methylated cytosine, but the cytosine on the complementary strand is unmethylated.

There an enzyme, Dnmt1, that specifically targets the unmethylated cytosine at a hemimethylated CpG and methylates it. So, after the action of this enzyme, the methylated state has been restored in each of the daughter cells. The combination of the hemimethylase (or maintenance methyltransferase) activity of Dnmt1 and the semiconservative replication of DNA set up a system in which the epigenetic state of an allele can be set once, and it will be propagated across multiple cell divisions.

DNA replication of a fully methylated CpG site results in two hemimethylated copies. The hemimethylase Dnmt1 then restores these hemimethylated sites to their fully methylated form.

In the next installment, we’ll talk about another epigenetic mechanism, histone modification, and the possibility of an analogous propagation mechanism for propagating those epigenetic marks.

Yoder, J., Soman, N., Verdine, G., & Bestor, T. (1997). DNA (cytosine-5)-methyltransferases in mouse cells and tissues. Studies with a mechanism-based probe. Journal of Molecular Biology, 270 (3), 385-395 DOI: 10.1006/jmbi.1997.1125

Genomic Imprinting V: DNA methylation and gene silencing

So, we’ve already discussed the fact that genomic imprinting is mediated through epigenetic differences between the maternally and paternally inherited gene copies. That is, at an imprinted locus, the maternally inherited allele will have one pattern of epigenetic modifications, while the paternally inherited allele has a different pattern. These differences are first established in the male and female germ lines, when the alleles that will eventually become maternally and paternally derived are in physically different locations. It is not hard to imagine, then, how these differences could be established. One pattern of gene expression in spermatogenesis results in the paternal-specific epigenetic modifications. A different pattern of gene expression in oogenesis results in maternal-specific epigenetic modifications.

But what are these epigenetic modifications, and how do they change the expression pattern of the gene?

There are a number of modifications involved in imprinting, but for the moment, we’re going to focus specifically on the simplest and best-understood mechanism: DNA methylation.

The two horizontal lines in this picture represent the two copies of a gene.  The big, solid box is the part of the gene that actually codes for the protein. The open box is the promoter region, which is the part of the DNA sequence responsible for regulating expression of the gene. The lollipop things indicate DNA methylation on cytosine residues (the “C” of the A, C, G, T alphabet that makes up DNA).

In this simplest type of scenario, the DNA sequence in the promoter region binds to a variety of proteins that recruit the molecular machinery that will transcribe the gene, leading eventually to production of the corresponding protein. The addition of methyl groups to the DNA changes its binding properties, so that it no longer binds to this machinery, and that copy of the gene is not transcribed.

If you’re not a molecular biologist, you can think of it like this. The transcription machinery is a bit like a Xerox machine, and the gene is like the master copy of some document. The promoter region is like a lock that has to be unlocked before you can copy this particular document. There are a number of proteins called “transcription factors” that function like a key to this lock. These transcription factors fit nicely on the promoter region, unlocking the gene and resulting in the production of many copies of the gene product.

Adding methylation to the promoter region is a bit like squirting epoxy into the lock. The presence of the methyl groups actually changes the physical shape and chemical properties of the DNA. So, when you try to put the key in, it no longer fits right, and the gene can not be copied.

In the top part, we see the red transcription factor binding to the black promoter region, which will activate transcription from the gene. In the bottom part, methyl (CH3) groups have been chemically added to the promoter region, preventing binding, and thereby preventing transcription.

So, these relatively subtle chemical changes are able to completely alter the functional properties of the gene.

Next time, we’ll talk about how these methylation patterns are maintained through development, and how the two gene copies are able to maintain distinct epigenetic states across multiple rounds of cell division and DNA replication. Make sure to tune in, because it’s really slick!

The two references represent the first proposals that DNA methylation might be the thing that permits the stable transmission of patterns of gene expression across cell divisions.

Holliday, R., & Pugh, J. (1975). DNA modification mechanisms and gene activity during development Science, 187 (4173), 226-232 DOI: 10.1126/science.1111098

Riggs, A. (1975). X inactivation, differentiation, and DNA methylation Cytogenetic and Genome Research, 14 (1), 9-25 DOI: 10.1159/000130315

Genomic Imprinting IV: Escalation Between Loci

So, in the previous installment, we introduced the “Loudest Voice Prevails” principle, which describes the evolutionarily stable pattern of gene expression at an imprinted locus where there is an intragenomic conflict over the total level of gene expression. Basically, the allele that favors lower expression becomes transcriptionally silenced. Expression from the other allele (the “louder” voice) evolves to the level that maximizes its inclusive fitness. In this sense, the active allele at an imprinted locus “wins.”

But what is going to happen if we have a pair of imprinted genes that exert opposite effects on the phenotype? If we have a paternally expressed growth enhancer, it will evolve to bring the growth phenotype up to the paternal optimum. If we have a maternally expressed growth suppressor, it will evolve to bring the growth phenotype down to the maternal optimum. But what if we have both?

Well, intuitively, if there is conflict between maternally and paternally derived genes over the optimal growth phenotype, then the phenotype can’t simultaneously satisfy the paternal and maternal optima. One or the other (or both) of these genes will always be under selection to increase its gene expression level (or, equivalently, the activity or longevity of the gene product, etc.). Thus, these two opposing genes will become involved in a kind of arms race.

In the simplest possible model that we can write down, this arms race goes on indefinitely, with natural selection driving each of the genes towards infinite expression. Clearly, in a real biological situation, this will not be the case, and something will step in to bring this escalation to a halt. The questions then become: What stops the escalation? And, what does the system look like at its new, escalated, evolutionarily stable state?

To think about this, let’s return to our analogy from last time, where Pat and Chris are sharing an office, but disagree about what temperature the office should be kept at. Recall that genes are totally passive aggressive, so Pat and Chris don’t compromise or communicate. They just use the tools at their disposal to move the office closer to their preferred temperature. Pat wants the office at 71 degrees. Chris wants it at 70.

We saw that if Pat and Chris both have space heaters, eventually Chris’s space heater is off, while Pat’s holds the temperature at 71. On the other hand, if they both have air conditioners, Pat will turn his/her A/C off, and Chris will get to have the room at 70.

If each of them has a space heater and an air conditioner, we have an arms race on our hands. Whenever the temperature is below 71, Pat will turn up the space heater. Whenever it is above 70, Chris will turn up the air conditioner. In passive-aggressive-allele fashion, this will go back and forth until the space heater and air conditioner are both blasting away. In the absence of any constraints or side effects, it will go on until both are blasting away infinitely.

There are several ways that the escalation could stop, however, each of which has a biological analog.

     (1) Mechanical limitation. There will be some limit beyond which gene expression / activity can not increase. Once one of the genes reaches its limit, the other will win. Like if Pat’s Tufnel-brand space heater goes to eleven, Pat wins. Of course, this will depend on the mechanisms through which the two genes exert their influence. For instance, if Chris’s air conditioner is actually a combination air conditioner / food processor / exfoliator, Chris might have to turn it way way up to get the air conditioning equivalent of a little bit of space heating. Similarly, a gene product might perform multiple tasks, and this pleiotropy could limit its competitive ability in the arms race.

     (2) Production costs. One difference between the single-locus solution and the two-locus solution is the level of energy consumption. If Chris’s space heater is off, Pat’s holds the temperature at 71. If Chris’s air conditioner is maxed out at ten, Pat’s space heater (which goes to eleven, remember) holds the temperature at 71. The difference is that the second solution comes with a huge-ass electricity bill. Can this sort of cost actually halt the escalation? Maybe. This requires either that there are diminishing returns to increased escalation, or that there are accelerating costs to production (like utility rates where your thousandth kilowatt-hour costs more than your first one).

     (3) Intervention. In a real office, we might expect that the manager would come in and yell at Pat and Chris, telling them to turn down their space heater and air conditioner. Maybe the manage would mandate an office temperature of 70.5 degrees. Does this ever happen with genes? Could a consortium of unimprinted genes step in and stop the escalation? There is no evidence to my knowledge of such things happening in the context of genomic imprinting, but this type of intervention is thought to be responsible for meiotic sex-chromosome inactivation, where the autosomes all gang up and put the sex chromosomes in a headlock in order to prevent meiotic drive.

     (4) Side effects. What if turning up Pat’s space heater also makes the music louder in the office? What if Chris’s air conditioner draws so much power that it causes occasional brown-outs? This is the other way in which the escalation between imprinted genes might be self limiting. If we consider a monolithic “growth phenotype” in isolation, then each allele has a simple, monolithic optimum. But genes are seldom like that. A paternally expressed allele may benefit from increased expression due to the effect of that increased expression on growth. But what if that increased expression has other consequences, as well? Maybe those other effects are detrimental to the allele’s inclusive fitness. If those deleterious side effects outweigh the growth-related benefits, then natural selection will not drive further escalation.

In future installments, we’ll look at some specific examples of escalating genes. But first, we’ll step back and look at some of the other features and consequences of imprinted genes.

Kondoh, M., & Higashi, M. (2000). Reproductive Isolation Mechanism Resulting from Resolution of Intragenomic Conflict The American Naturalist, 156 (5), 511-518 DOI: 10.1086/303409

Wilkins, J., & Haig, D. (2001). Genomic imprinting of two antagonistic loci Proceedings of the Royal Society B: Biological Sciences, 268 (1479), 1861-1867 DOI: 10.1098/rspb.2001.1651

Genomic Imprinting III: The Loudest Voice Prevails

So, it’s been a while since the last installment of the Primers on Imprinting feature, but they should be posted with greater regularity in the upcoming weeks. This time we’re going to introduce something that we will see again in future installments: small differences in selection lead to large differences in behavior.

Last time, we introduced the most widely discussed and most successful explanation of the evolutionary origins of genomic imprinting, the “kinship” or “conflict” theory. According to this theory, imprinted gene expression is a consequence of the fact that natural selection acts differently on alleles depending on their parent of origin. There are several ways to think about the origin of this differential selection, but we talked about it in terms of the framework that I find most intuitive: inclusive fitness.

As we also noted last time, even in the cases where the asymmetry in selection on maternally and paternally derived alleles is sufficiently large to drive the evolution of imprinted gene expression, the actual magnitude of this asymmetry is actually incredibly small. Why? Well, even for a allele with large effects on the survival and reproduction of related individuals, the dominant factor in the inclusive fitness of that allele is still going to be the survival and reproduction of the individual organism carrying that allele around.

But, the standard pattern observed with imprinted genes is that the allele-specific expression is all or nothing. For example, an allele might be expressed when it is inherited from a male, but completely silent when inherited from a female. So this small difference in the optimal expression levels of the maternally and paternally derived alleles leads to – in a way – the largest possible difference in the realized expression levels of the two alleles.

I like to think of this in terms of an analogy. Imagine that Pat and Chris share an office, and that they have a slight disagreement over the temperature they want the office at. Say Pat wants the office to be at 71 degrees (Fahrenheit), while Chris wants it to be 70. Each of them has control over a small space heater, and this is the only mechanism that they have for manipulating the temperature. [1]

What’s going to happen? Let’s say the temperature is 70 degrees. Pat will turn up his/her space heater until the temperature reaches 71. In response, Chris will turn his/her space heater down until the temperature comes down to 70. They will go back and forth like this until, eventually, Chris’s space heater is completely turned off. Pat will then turn his/her space heater up to get the room to 71. Then we’re done. Chris is unhappy about the temperature of the room, but no longer has any ability to make it any cooler.

Two things about this outcome. First, a small disagreement over the ideal temperature has led to a large divergence in the strategies: Chris’s space heater is all the way off, while Pat’s is on and doing all the work. Notice that the outcome would be exactly the same, in principle, if Chris’s ideal temperature were 70.9 degrees, or even 70.999 degrees. [2]

Second, Pat wins. This is a consequence of the fact that we are talking about space heaters, and that Pat prefers the higher temperature. If, instead of space heaters, Pat and Chris each had control of an air conditioner, Chris would be the winner. At equilibrium, Pat’s air conditioner would be all the way off, and the room would be at 70 degrees.

This is also the way it works with alleles at an imprinted locus. Let’s consider the case of a gene where increased expression results in increased prenatal growth. The inclusive fitness argument says that the optimal amount of this growth factor is higher for an allele when it is paternally inherited than when it is maternally inherited. Say this patrilineal optimum is 105 units, while the maternal optimum is 95 units.

If the gene is not imprinted we might expect it to produce about 100 units, with each allele producing 50. However, once the evolutionary dynamics of imprinting take over, the pattern of expression will evolve to one where alleles are transcriptionally silent when maternally inherited, but where a paternally expressed allele is making 105 units.

For a growth-suppressing gene, where increased expression actually reduces prenatal growth, we expect the opposite pattern, where alleles are silenced when paternally inherited, but are expressed when maternally inherited. This set of predictions – that imprinted growth enhancers will be paternally expressed, and imprinted growth suppressors will be maternally expressed – matches the empirically observed pattern by and large, although there are a few counterexamples that are not fully understood at the moment.

This pattern of allele silencing has been dubbed the “loudest voice prevails” principle. The phenotype evolves to the optimum of the allele favoring higher expression. Now, you can argue that this is the sort of thing that does not really need its own name. Fair enough. It’s really just saying that the evolutionarily stable state of the system is an edge solution. But, “loudest voice prevails” is sort of catchy, and has the advantage of reminding us which allele is expressed at equilibrium.

The Haig 1996 citation is the paper that introduces the phrase. The other three citations are papers published around the same time that use different mathematical frameworks to address the evolution of gene expression at an imprinted locus. Generically speaking, the answer is the one described here, although the Spencer, Feldman, and Clark paper identifies certain regimes in parameter space where apparently different results can be obtained. In a future post, we will delve into the differences in the assumptions and conclusions of different modeling frameworks as they have been applied to imprinting.

Now, what if you consider more than one imprinted gene? What if Pat and Chris each have a space heater and an air conditioner? We’ll talk about that next time.

Haig, D. (1996). Placental hormones, genomic imprinting, and maternal-fetal communication Journal of Evolutionary Biology, 9 (3), 357-380 DOI: 10.1046/j.1420-9101.1996.9030357.x

Mochizuki A, Takeda Y, & Iwasa Y (1996). The evolution of genomic imprinting. Genetics, 144 (3), 1283-95 PMID: 8913768

Haig, D. (1997). Parental antagonism, relatedness asymmetries, and genomic imprinting Proceedings of the Royal Society B: Biological Sciences, 264 (1388), 1657-1662 DOI: 10.1098/rspb.1997.0230

Spencer HG, Feldman MW, & Clark AG (1998). Genetic conflicts, multiple paternity and the evolution of genomic imprinting. Genetics, 148 (2), 893-904 PMID: 9504935


[1] Of course, in the real-life situation, we might assume that Pat and Chris would discuss the situation and come to some sort of agreement. This is a key difference between people interacting in strategic situations and genes evolving under natural selection. Alleles at a locus are like people sharing an office, where both of them are incredibly passive aggressive. If it helps, imagine that Pat and Chris won’t talk to each other.

[2] In practice, of course, there is going to be some minimum level of disagreement required in order to trigger this passive-aggressive escalation. In this analogy, the minimum level will be set by a combination of things such as the sensitivity of Pat and Chris to small changes in temperature, the precision with which the space heaters control the temperature of the room, and the extent to which they care about each other’s comfort. Similar reasoning holds in the case of genes, and we will address this in a future installment of the series, where we ask why there are any genes that are not imprinted.

Genomic Imprinting II: Inclusive Fitness and Conflict

So, as we discussed previously, genomic imprinting is the phenomenon where the pattern of expression of an allele differs depending on whether that allele was inherited from your mother or your father. This difference in expression does not depend on differences in the DNA sequence of the two alleles. Your two alleles might have identical DNA sequences, but function completely differently.

For example, one of the two alleles might be epigenetically silenced. That is, because of reversible chemical modifications to the DNA, or to proteins that are closely associated with the DNA, that allele would be inactive. The other allele, with an identical DNA sequence, but different modifications, would be happily chugging along producing its gene product(s). Today’s question is, “That seems crazy! Why would you do something like that?”

One of interesting (and by “interesting” I mean “sad”) things about evolutionary biologists is that whenever something genuinely new and surprising is found in the world, everyone feels the need to propose an evolutionary explanation for it, whether new explanations are needed or not. So, many attempts to explain the origins of imprinting have been proposed, most of which are consistent with at least some of the data, and a few of which actually make sense. There is one explanation, however, that is far and away the most successful in explaining the distribution and nature of imprinted genes: the “Kinship” or “conflict” theory of imprinting. This theory owes its creation and early development primarily to David Haig (who laid out the theory originally in papers with Mark Westoby and Tom Moore).

The basic idea of the Kinship Theory is that natural selection favors different expression behaviors for maternally and paternally inherited alleles. That is, the optimal level of expression for an allele that is maternally inherited is different from the optimal level for a paternally inherited allele. Now, if you think about that for a minute, it probably seems strange. I mean, if I survive and reproduce, that’s equally good for any of my genes, independent of where I got them from, right? Right?

The key is recognizing that natural selection favors allele that pass on the largest number of copies to future generations. (“Isn’t that what I just said?”) AND, that it doesn’t matter whether those copies are passed on directly by my having children, or if they are passed on because because one of my relatives, who has inherited an identical copy of that allele, has children. There are different ways to represent this (which are all mathematically equivalent if you do them right), but the one that I find most intuitive is the idea of “inclusive fitness.” Basically, we can think of natural selection as maximizing the inclusive fitness of an allele, which is the fitness of every individual in the population, weighted by the probability that they carry the allele. In the simplest model, this probability is 1 for me, 1/2 for a sibling, 1/8 for a cousin, and so on.

Now, think about your relatives. With the exception of your descendants, your full siblings, and their descendants, all of your relatives are related to you either through your mother or your father. That means that, in general, this inclusive fitness calculation will be different for maternally and paternally derived alleles. Therefore, there will be a conflict, where the phenotype that would maximize the inclusive fitness of your maternally inherited allele will be different from the one that would maximize the inclusive fitness of your paternally inherited allele. If this difference is large enough, it can actually drive alleles to take on two different conditional expression strategies, and, voila, imprinting.

Granted, even in the most extreme cases, this difference is likely to be pretty subtle, which might make it seem strange that this could explain a situation where one of the two alleles is completely turned off, while the other one is cranking away. This phenomenon, where a little conflict leads to a big effect, will be the subject of the next installment.

Genomic Imprinting I

So, one of the things that I study is genomic imprinting. What is that, exactly?

Even if you’re not a biologist, you are probably familiar with the fact that, for most of your genes, we carry two different copies, or alleles. You get one of those alleles from your mom, and one from your dad. Those two alleles could be the same (have identical DNA sequences) or different (usually only at a small number of positions within DNA sequence). If they are different, then the consequences of those alleles on your traits, like how tall you are or what color your eyes are, are determined by the dominance relationship between those two alleles. For example, the main allele responsible for red hair (at the MC1R locus) is recessive in relationship to alleles for brown or black hair. So, if you have only one copy of the red-hair allele, you will probably have dark hair. Importantly, in terms of what follows, it does not depend whether the recessive red-hair allele you have came from your mother or father.

If you are a biologist, you already knew all of that, but you may or may not be familiar with imprinted genes. About one percent (or possibly more) of our genes are imprinted. For these genes, it does matter which allele came from your mother and which one came from your father. That’s because imprinted genes retain a chemical memory of which parent they came from, and function differently depending on their parental origin. More specifically, at an imprinted locus, alleles are subjected to epigenetic modifications in the germ lines (ovaries or testes). These epigenetic modifications can be chemical modifications applied directly to the DNA itself, or modifications to proteins that are closely associated with the DNA. These modifications alter how the allele functions, without modifying the DNA sequence itself. The key thing is that, for imprinted genes, the epigenetic modifications that are established in the male germ line are different from those established in the female germ line. So the allele that came from your father will function differently from the allele that came from your mother, even if the DNA sequences are identical.

In the simplest cases, one of the two alleles is inactivated, or turned off. The effect of that gene on a given trait, then, depends only on the active allele. To return to the red-hair example, imagine that the MC1R locus was imprinted (which it is not, as far as we know), and that only the paternally inherited copy was expressed. Now, if you had one copy of the red-hair allele, and one of the more common dark hair allele, you would not necessarily have dark hair. Your hair would be dark if your red-hair allele came from your mom, but if it came from your dad, your hair would be red.

Of course, as with all things in biology, once you start looking at the details, everything becomes a lot messier and more confusing. But, that is the basic gist.

Genomic imprinting was one of the biggest surprises to come out of molecular biology in the past few decades. Both the origins of imprinting of particular genes, and the effect of imprinting on the evolution of those genes, are interesting questions that we will return to in future posts. At some point along the way, we will get deeper into those messy and confusing details.