Last week I began a breakdown of this paper. It's a much more complicated paper than I usually cover round here, and I will also be covering it in more depth than usual, because I think there are a lot of things about it that are worth discussion, and I think that even this kind of complicated paper, requiring a ton of background knowledge, is perfectly understandable if I can give you that background knowledge (amusingly, the abstract of all scientific papers is supposed to be "understandable by the lay public", as well as being often under 200 words. Yeah, like THAT'S gonna happen). So, let's get in to part 2!!
Last time, I gave a basic background into DNA methylation, what it is, and how it relates to this paper. Please go to the other post to get more background on that one, but VERY basically, DNA methylation controls whether a particular section of DNA encoding a gene is "available" to make proteins. The proteins that CONTROL DNA methylation are DNA methyltransferases (Dnmt). Last time, the first figure of the paper showed that the expression of Dnmt3a (the 3a DNA methyltransferase) was increased (and then decreased) following injection of a SINGLE INJECTION of cocaine, and was increased as well at 7 injections, or 28 days worth of cocaine treatment or self-administration.
But now that we know that Dnmt3a is changed, the next question is, what is the EFFECT of that change? How does it change the way the animal responds to cocaine? How does it change behavior?
To look at this, they went from several different angles. First, they gave an series of injections that increased DNA methylation across the board. The idea behind this is to mimic the possible effects of increase Dmnt3a that they saw in animals who got cocaine, but in this case they want to look at JUST the methylation effect, and not any other effects of cocaine. So you can't give cocaine, instead you have to increase DNA methylation. This would be a major issue if they did it all over the brain, so they did it instead in a very specific area, the nucleus accumbens. This is an area of the brain that plays a large role in the rewarding and reinforcing effects of cocaine and other drugs and things that are rewarding.
You can see up there the nucleus accumbens (where they are referring to is the mouse brain, looking at it from the front, and sliced about halfway back, as you'd slice a loaf of bread).
So they increased DNA methylation just in the area of the nucleus accumbens, and then they looked to see how this affected the rewarding properties of cocaine in mice. They did this using a paradigm called conditioned place preference.
Take a box, and divide it into two chambers (though it's often done with three, two will work for this explanation). The two chambers on either side are decorated differently. Maybe they have different floors, or different stripes or spots. Or even different smells. For this, say you have a pink chamber and a blue chamber.
You give the mouse an injection of cocaine and plunk it in the pink chamber.
Then you wait a while (for the cocaine to wear off), and give the mouse an injection of saline, and plunk it in the BLUE chamber.
You continue this for a few days, always pairing the pink chamber with cocaine and the blue chamber with saline.
Then, on the final day, put the mouse in with no injections, and see where he spends his time.
If the mouse found the cocaine rewarding (as you might expect he would), he'll spend more time in the pink chamber, whereas if he found it aversive, he would spend more time away from the drug chamber and in the blue side. This test is used for all sort of drugs and can be pretty sensitive.
SO. They increased DNA methylation in the nucleus accumbens, and then looked to see how that affected the conditioned place preference of the mouse for cocaine.
Let's start with the graph on the left. That's the increase in DNA methylation, and they got a DECREASE in cocaine conditioned place preference, meaning the mice "liked" the cocaine LESS when there was more DNA methylation going on.
To see if this worked both ways, they also gave a drug to decrease DNA methylation (that's the center graph up there), and ran the same experiment, and this time found an INCREASE in conditioned place preference. So mice liked cocaine MORE when there was LESS DNA methylation.
They also looked at locomotor sensitization, which is a test where you give several doses of cocaine. Cocaine is a stimulant, and increases locomotor activity (if you've ever seen anyone on coke, you'll know what I mean). If you give several repeated injections to a mouse, you will get what's called sensitization, where the locomotor activity increases even more than it did at first.
What you can see here is locomotor sensitization, and it appears that DECREASING DNA methylation, which INCREASED conditioned place preference, also increased cocaine sensitization, so that all lines up.
Now, let's get back to Dnmt3a. First off, they used a viral vector to increase Dnmt3a in the nucleus accumbens only. This is a cool newish technique, where you take a virus (in this case they used HSV), take out the actual nasty bits, and put IN DNA which codes for your gene of interest. In this case, Dnmt3a. When you inject the virus, the virus will infect the cells, but instead of causing the production of nasty virus, it causes the production of YOUR GENE. This is a very powerful tool for us scientists and is also pretty cool to explain. We have mastered the virus for our own nefarious purposes! MWAH HA HA HA!!!
(Mine is an EVIL LAUGH!)
So anyway, they increased Dnmt3a in the nucleus accumbens. Previously, they found that Dnmt3a was increased in response to cocaine injection. So what happened when they increased Dnmt3a and then looked at conditioned place preference?
This is where it gets a little crazy to me. Increasing Dnmt3a only in the nucleus accumbens DECREASED conditioned place preference, it made it LESS rewarding. Conversely, they also found they if they selectively knocked out Dnmt3a in the nucleus accumbens (using something we call floxed mice, which I will have to explain later), they got INCREASES in conditioned place preference.
So the authors concluded that increased DNA methylation decreases cocaine reward, and decreased DNA methylation increases cocaine reward.
And I'm a little confused by these findings. Considering the previous figure, where we saw persistent increases in Dnmt3a (and thus DNA methylation) following cocaine self-administration, why are they seeing DECREASED sensitization and conditioned place preference? The sensitization is harder to define (what it actually means is still a mystery), but if the animals are "addicted" to cocaine based on the long term cocaine administration, and that results in increased Dnmt3a over a long period of time, WHY does increased Dnmt3a result in decreased REWARD from cocaine?
If I had to guess why this was happening, I would guess for something like tolerance, but I'm also surprised by the fact that they got OPPOSITE effects when they decreased DNA methylation. The authors didn't really cover this in the discussion, just said that higher Dnmt3a decreases cocaine reward, while lower Dnmt3a increases cocaine reward. But based on their results from their first figure, I think it may be more complicated than that.
And now the question is, what could they have tried to address this problem? They could try looking with the rats that had been self-administering cocaine for 28 days, and had an increase in Dnmt3a, and seen how they did in conditioned place preference. Does that result in decreases like they saw with the mice? What about things like motivation for cocaine? Do rats have higher progressive ratios (a measure of how hard rats will work for a dose) when they have higher levels of Dnmt3a? As it is, I don't feel like the results line up with each other.
But don't worry, we're not done yet! Next up, spine density and...depression?
LaPlant Q, Vialou V, Covington HE 3rd, Dumitriu D, Feng J, Warren BL, Maze I, Dietz DM, Watts EL, Iñiguez SD, Koo JW, Mouzon E, Renthal W, Hollis F, Wang H, Noonan MA, Ren Y, Eisch AJ, Bolaños CA, Kabbaj M, Xiao G, Neve RL, Hurd YL, Oosting RS, Fan G, Morrison JH, & Nestler EJ (2010). Dnmt3a regulates emotional behavior and spine plasticity in the nucleus accumbens. Nature neuroscience, 13 (9), 1137-43 PMID: 20729844