Saturday, November 29, 2014

Improving reproductive cloning

Remember when Dolly the sheep was cloned in 1996?  That was the first cloned mammal and everyone freaked out thinking we would be cloning all our pets and even humans within a few years.  Well, nearly 20 years have passed since then and reproductive cloning is still a very difficult and inefficient procedure.  Most cloning has a 1-5% success rate.  Why is that?  Before we can answer that, we need to understand the procedure for reproductive cloning.

Somatic cell nuclear transfer
Our bodies are made up lots of different types of cells – neurons, skeletal muscle, intestinal cells, immune cells, etc.  Despite the different functions and structures of these cells, all of the cells in one organism have the same genome, the same set of genes.  What makes cells unique is that they express different genes at different times, so different proteins are made.  What this means is that the blueprint (DNA) for making a new organism is right there in every cell in your body.

The normal way of making an embryo is by taking half of the genome from a male (in sperm) and half from a female (in the egg) and combining them during fertilization.  In reproductive cloning, you already have a whole genome from any adult cell.  That nucleus from the adult can be inserted into an oocyte (or egg) that has had its DNA removed (“enucleated”).  The egg is necessary because it has lots of nutrients and signals in it that are important for the first few cell divisions during early development.

This process is shown in the diagram below and is called somatic cell nuclear transfer.  In the example, the adult genome is coming from a fibroblast cell and is transferred into an enucleated oocyte.  This is a way to get stem cells (ntES), which can be used for therapeutic purposes, like making more neurons that can be transplanted into someone with Parkinson’s.  Or you could let the cloned cells grow up into an embryo and then into a cloned organism.

Somatic nuclear cell transfer (from

Although it is possible to make cloned organisms using adult donor genomes, the efficiency is much higher when using genomes from embryos.  What happens to the adult genome that prevents it from directing the formation of a new organism?  This problem is addressed in a recent paper by Matoba et al., published in Cell.

Epigenetic changes
Although adult cells should have the same DNA sequences as their embryonic precursors, the genome can be organized differently, which can affect which genes are expressed.  DNA wraps around histone proteins as a way to organize the long DNA chains.  Histones can be modified in such a way that the DNA will wrap around more tightly or more loosely.  For example, if a particular amino acid in histone 3 is trimethylated (three CH3 groups are added), then that makes the DNA pack up closer together, so it is really hard to express those genes.  There are genes that may need to be expressed early on in development, so their histones will be modified to allow for loose packing, but then after they are expressed, they’ll get packed away, so they take up less space.  These kinds of modifications that affect gene expression are called epigenetics.

As an analogy, imagine you have had a child, so you have baby clothes, a crib, car seat and toys in your house.  Once that child grows up, you take all those baby things down to the basement.  You still have them, but you will probably never need to use them again, so you can pack them all up and store them so they are out of the way.  It may be hard to access them again, but they are still there.  So a gene that has been packed away into condensed chromatin is still present in a cell, but it is no longer giving instructions for making proteins, unless something comes along and unpacks it.

You can see now the problem with somatic cell nuclear cloning.  The adult cell already has some DNA packed away, so when the genome is put into an oocyte, it may be impossible to express the genes necessary to direct normal development.  In the paper by Matoba et al., they did indeed find that there are more trimethyl modifications (called H3K9me3) in mouse embryos derived from nuclear transfer than embryos from in vitro fertilization (using a sperm and egg).  These regions were associated with decreased gene expression and compact DNA.

Improving efficiency
Now we know one of the problems, but what can researchers do to improve the efficiency of reproductive cloning?  Somehow they need to decrease H3K9me3 modifications in the donor genome.  They do this two ways:

(1) There are too many methylated histones, so the authors injected an enzyme that removes methyl groups into the one-cell embryos.  The embryos expressed more genes and survived throughout development.  70% of these cloned embryos implanted into a surrogate mouse uterus and 8% survived to adulthood.  Those numbers are higher than before, but still not perfect.

(2) They also tried decreasing expression of the enzymes that put on the methyl groups.  This also improved development, so 50% of the embryos made it to later stages of development, but they did not see how many survived to adulthood.

There is an epigenetic barrier for nuclear transfer from adult cells in mouse oocytes.  Presumably a similar problem is preventing cloning in other organisms as well.  It makes sense that an adult cell would have a different pattern of epigenetic modifications than an embryonic genome.  The authors were able to improve cloning efficiency by decreasing the H3K9me3 modification, but there are probably other histone modifications that are also different in adults.  There is still a long way to go before cloning is a reliable procedure, but at least now we have some explanation of why it is so difficult.

Sunday, September 7, 2014

Transgenerational inheritance of fear

A new semester has begun and I have no extra time to update this blog, so just a short entry today.  This paper was just too cool to pass up.  It was published earlier this year in Nature Neuroscience by Dias and Ressler.  They conducted a series of experiments which showed that learned fear can be passed on from generation to generation in the sperm DNA.  That’s Lamarckian evolution for all you evolution nerds out there.  A learned behavior that is inherited genetically -- totally crazy!

The DNA sequence itself isn’t changing, but instead the expression of genes is altered, so different amounts of proteins are being made.  This process is known as epigenetics (which I’ve discussed before with regard to histone modifications).  One way to change DNA expression is by methylating cytosines (the “C” in DNA sequences).  The methyl group (CH3) makes it harder for proteins to bind to the DNA and transcribe the genes into mRNA and subsequently into protein.  The general rule of thumb is: more methylation --> less gene expression, less methylation --> more gene expression.  This is a common way our cells regulate gene expression, and what’s really interesting is that many external influences can affect DNA methylation, like traumatic life experiences, smoking, exercise, environmental toxins, etc. 
Cytosine getting methylated (note the H3C addition on the the molecule on the right)

It is conceivable that DNA methylation patterns can be inherited through generations, so changes in gene expression that affected your great grandparents could still be maintained in your cells.  Most of these types of studies focus on changes that occur to the mom during pregnancy.  For example, lets say that researchers expose a pregnant mouse to a toxin that may affect DNA methylation.  The next generation (the F1 generation) is also being exposed in utero to the same toxin.  The eggs or sperm progenitor cells are also developing in the embryo, so the next next generation (F2) may also be exposed to the toxin.  Thus, if you see DNA methylation changes in the F1 and F2 offspring, this isn’t really so surprising since these cells were all exposed at the same time as the mom (F0 generation).  In order to really prove that DNA methylation patterns can be inherited across generations (transgenerationallly), you need to expose the parents to a stimulus before conception, before the F1 and F2 generation even exist, which is what they did in this paper.

The authors chose to initiate changes to gene expression by conducting odor conditioning in mice.  They paired a particular odor with a mild foot shock and conditioned F0 males to be afraid of the odor.  Then these males mated with naïve females (never exposed to the odor).  The F1 offspring showed excessive fear to the conditioned odor, even though they had never encountered it before.  It was their fathers who had been shocked, not them.  The authors found there were more cells in the olfactory region of their brains that expressed the olfactory receptor for the conditioned odor.  Furthermore, their dad’s sperm and their own sperm were less methylated in the gene that encodes for that particular odor receptor, so the offspring of F1 were also affected.

F0 dad learns fear for odor --> decreased methylation for receptor gene --> F1 offspring inherit methylation changes --> express more of the receptor --> increased sensitivity to odor --> F1 sperm have same decreased methylation --> sensitivity for odor passed on to F2 generation  

Grandpa (F0) had a bad experience with an odor and now his grandkids will be more afraid and sensitive to that odor.  And it’s all genetic.  The authors did a series of experiments to show that it isn’t behaviorally based (grandpa isn’t telling the grandkids about his horrible experience with this smell).  For instance, they took the sperm from the F0 mice and took it to a different mouse facility and did in vitro fertilization, so the father was nowhere near his offspring or the mother.  The F1 mice from in vitro fertilization were just as super sensitive to the odor.

This is nuts!

Saturday, May 24, 2014

Ctenophores come before

Three months ago, if I had seen this article about the ctenophore genome, I would have moved right passed it without a second look.  What is a ctenophore and why would I care about the sequence of its DNA?  But then I taught Bio 2 this spring and learned about animal diversity and the evolutionary tree (a day before I taught it).  This is a great example that the more you learn, the more interested you become in the subject.  Today’s article by Moroz et al. was published recently online in Nature (this one is open access, so take a look).  The results totally change the roots of the animal tree and invalidate what we taught to our students this semester.  Before we get into the paper, let me answer my own questions:

What is a ctenophore?
Ctenophores are also known as comb jellies, because they look like jellyfish and have a comb-like structure that they wiggle around to move through the water.  They have sensory organs to sense light and gravity.  They have tentacles that they move with their nervous system in order to catch prey.
Comb jelly (from Wikimedia Commons)
Although they look like jellyfish, they are in a totally different phylum, or branch of the evolutionary tree (also known as a phylogenetic tree).  Jellyfish are in the phylum Cnidaria, along with sea anemones, coral and hydras.  Comb jellies are in their own phylum known as Ctenophora. 

Why should we care about ctenophores? 
Ctenophores, along with cnidarians and sponges, represent some of the most ancient lineages of animals.  Studying them can give us a clue about how animals evolved.  All the rest of the animals are in the large taxonomic group called Bilateria, because they have bilateral symmetry (which is symmetry across a single axis).  This includes humans, insects, crustaceans, worms, fish, molluscs, etc.  Think about a jellyfish or a sea anemone; they have radial symmetry, which means they can be bisected in lots of different axes and you would still have symmetrical halves.  The bilateral animals are more complicated in lots of other ways, such as having a greater variety of tissues and more complex physiology.

The phylogenetic tree according to the biology textbook
Sponges don’t have organized tissues and they don’t have a nervous system, so based on that, researchers have considered them to be the most ancient lineage (i.e. the “basal” animals).  So if we are building a phylogenetic tree based on morphological characteristics, we are going to put them as the first branch.

Cnidarians and ctenophores look very similar, so it would make sense to put them right next to each other, followed by the bilateral animals.  Thus, based on morphological observations, the tree should look something like this:

No one ever explained to me why cnidarians get to be closer to the bilaterals than ctenophores.  Perhaps this is because the cnidarians come in so many different body plans, so maybe they are considered to be more complex and thus, an evolutionarily “newer” animal.

We told our students over and over: “Sponges are the most basal animals”.  But like many phylogenetic theories that have come before, new DNA sequencing data is challenging this view.

What does the ctenophore genome tell us?
First off, for the non-biologists out there, you need to understand one fundamental thing about gene evolution.  Two species that are highly related will have very similar DNA sequences.  The further apart two species are in evolutionary time, the more time there is for mutations to change the DNA sequences and the gene functions.

Moroz et al. sequenced one of the ctenophore genomes and then compared it with the genomes of sponges and cnidarians.  One of the main findings was that there are many missing animal-specific genes that are involved in animal development (Hox genes), regulating gene expression (no miRNAs!) and innate immunity.  Although some animal-specific genes are absent, the ctenophores have many unique genes that are not found in other animals, indicating that these genes evolved independently in the ctenophore lineage. 

The researchers devoted a lot of the paper to looking at genes involved in nervous system function.  Ctenophores, like cnidarians, have neural nets, as opposed to organized bundles of nerves.  Bilateral animals have many different neurotransmitters, which are the signals that get sent between nerve cells.  The ctenophores only have genes for making the neurotransmitter glutamate (and GABA), but they have a ton of glutamate receptors, more than other animals.

All of these findings led the authors to conclude that ctenophores are the most basal animals, not sponges.  Alternatively, it is still possible to keep the same phylogenetic tree, but there would need to have been massive gene loss in the ctenophore lineage.  The most parsimonious explanation is shown here:

The main difference between these two trees is that the sponges and ctenophores have swapped positions.  Note that this would require a nervous system to have developed twice independently.  That’s totally insane.  The ctenophores and the cnidarians “needed” a method of controlling their body to capture prey and both lineages “came up” with the same solution (of course evolution is random and doesn’t have a particular goal in mind).  When similar structures evolve independently, this is known as convergent evolution.

Next year, instead of devoting half a lecture to sponges and tossing in a single slide on ctenophores, I think I’ll have to give ctenophores their due, as potentially the most ancient lineage of animals still in existence. 

Carl Zimmer always beats me to the punch, so here’s his take on the same article.  Better writing, but fewer trees!

Wednesday, January 15, 2014

The mosaic female brain

Female mammals have two copies of the X chromosome while males have only one copy (because they have a Y chromosome instead).  Chromosomes contain genes and genes are the instructions for making proteins, so if females have twice as many copies of each gene on the X chromosome, will they make twice as much protein?  The answer to that is mostly “no”.  In young female embryos, one X chromosome is randomly inactivated and will remain that way through her life.  The chromosome gets compacted into a structure known as a Barr body.  However, when X inactivation occurs there are many embryonic cells and each one can inactivate one copy or the other.  Why does this matter?  Well, remember that one X chromosome came from dad and one from mom, so there may be different variants for each gene; different versions of proteins can be made depending on which X chromosome is still active in that cell.  In other words, females are genetic mosaics, where each cell may express one X chromosome or the other.  That’s cool!

What if the female embryo inherits one good copy of a gene and one bad copy that is non-functional and disease-causing?  Some of her cells would express the good copy of the gene and be fine and other cells would express the bad copy and be messed up.  The severity of the disease for this female will depend on how many cells inactivated the good copy and where these cells are located in the body.  Imagine that X inactivation occurred at the 4 cell stage, where two cells inactivate the good chromosome and the other two cells inactivate the bad chromosome.  Once an X chromosome is inactivated, it will stay that way in all the cells that are formed from that original cell in the 4 cell stage (see the figure below).  If each one of those 4 cells divides the same amount to form the final adult form, then you would expect half of her cells to be messed up and half of them to be fine.  But what if the two cells with the active bad chromosome happen to be cells that will divide way more and make way more future tissues of the body?  Then in the adult form, she would have tons of messed up cells and probably have a much more severe version of the disease.

In females one X chromosome is inactivated early in development (image from

As I mentioned earlier, X inactivation actually happens later on in embryonic development when there are more cells and each one can choose to inactivate one chromosome or the other.  If we consider the disease scenario, this random nature of X inactivation can lead to huge variability in X-linked disease expression in females.  It’s also important to think about how certain types of cells and tissues develop.  If an entire tissue develops from a single cell after X inactivation, then all of the cells in that tissue will have the same inactivated chromosome. 

Researchers at John Hopkins University visualized X inactivation by marking expression from each X chromosome with a different fluorescent protein.  Wu et al published their beautiful images in a recent article in the journal Neuron.

Marking X chromosomes
The authors created two types of mice, which each had an extra inserted gene on the X chromosome.  One type had a gene that encodes a red fluorescent protein called tdTomato.  The other mice had a gene for the green fluorescent protein, or GFP, which was originally discovered in jellyfish.  They then mated these two mice together and used the female offspring that had one X with tdTomato (Xt) and one X with GFP (XG).  If the “red” chromosome is inactivated, then only GFP will be expressed and this cell will look green, as will all of its daughter cells.  This way they can look at the heterogeneity of X chromosome expression in different parts of the body. 

The results
Overall, the mice came in all different amounts of red and green.  For instance one mouse might be nearly all green while its sibling is all red, again indicating that X inactivation is a random process.  In the mice that had both red and green, it was interesting to see the different patterns in the body.  For instance, in the intestine, cells of the same color were found in columns.  That’s because the cells in the column originate from one single stem cell, so they should all contain the same active X chromosome.

X inactivation appears in columns in intestinal tissue, because cells from a single stem cell migrate together

 Another interesting finding was that skeletal muscle cells expressed both red and green fluorescent proteins.  This would seem to indicate that there is no X inactivation in muscle, but this is not the case.  Skeletal muscles are actually formed by muscle progenitor cells (myoblasts) that fuse together, creating cells with multiple nuclei and copies of the genome.  If a cell with an active “green” X chromosome and a cell with an active “red” chromosome fuse together, then the muscle will express both proteins.  This only works for skeletal muscle; cardiac muscle in the heart does not develop by cell fusion, so these muscle cells are either red or green.  This is a great demonstration of the differences in muscle development.

Skeletal muscle cells express both X chromosomes, because they are formed via cell fusion

 They also noticed clear differences between the left and right side of the body, like in the tongue, retinas and brain.  This indicates that progenitor cells stay segregated to either the left or right side during development.  In other words, there is not a lot of migration between the two sides of the body, where a cell on the right side would make cells for the left side of the body, and vice versa.

The mosaic brain
The main focus of this paper is on the heterogeneity in the nervous system.  They looked at two different cell types in the brain: excitatory pyramidal cells and inhibitory interneurons.  These two types of neurons develop from different areas of the embryonic brain.  They found that inhibitory interneurons were highly mixed.  When they quantified the fraction of red inhibitory cells in two different parts of the brain, the values were very similar.  On the other hand, when looking at excitatory neurons, there was a lot of variability of which X chromosome was inactivated, across different parts of the brain and in different animals.  If there was an X-linked gene that affected excitatory neuron function, then the effects on neuronal circuits would be different for different regions of the brain in an individual.  The authors suggest that this could actually be a good thing, because it would allow females with different genetic variants to respond to a range of stimuli, increasing the dynamic range. 

So there are bad aspects of X chromosome inactivation, like the expression of X-linked diseases, but there are also some good points, like increased functional diversity of neurons.  The authors suggest that X inactivation “may represent one of the more significant mechanisms by which individual differences in central nervous system function are generated.”  It is crazy to think that random inactivation of a chromosome in the early embryo might give us our future individual personalities.