When scientists think of the pathways in the brain that underlie the responses to drugs, especially reward related responses, we tend to think of certain things. We think of the neurotransmitter dopamine, and we think of the nucleus accumbens. The nucleus accumbens is a bunch of neurons lying about three inches in from your ears (one bunch on each side, of course), and it plays an important role in things like the feeling of and expectation of reward. This means that there is a lot of activity in this area in response to things like drugs, which, you know, pack a LOT of reward.

While many of the papers that I read which talk about drug reward and addiction related processes focus on the nucleus accumbens and on the dopamine which spikes there (which feeds on to GABA neurons which make up the majority of the neurons in the accumbens), this paper is different, and as such, is pretty eye opening, to me at least. This paper focuses on GLIA. And it focuses on something other than dopamine...it focuses on cytokines.

Schwarz et al. "Early-life experience decreases drug-induced reinstatement of morphine CPP in adulthood via microglial-specific epigenetic programming of anti-inflammatory IL-10 expression" Journal of Neuroscience, 2011.

We'll start with glia. The glia are the understudied cousins of famous neuron. I would say they're the 99%, but they're only slightly more than half of your brain cells total (though they are the majority). But very few people study glia (compared to the number of people who study neurons, anyway). Even in grad school, I was told that glia wrap around neurons, making it easier to send action potentials, maybe send some nutrients or take up stuff...but meh, who cared about glia? But it turns out that glia may be much more important than we previously thought. For one thing, glia are the main movers in the immune response in the brain.

And here is where we get into cytokines. When cells that participate in the immune response are stimulated (by stress, injury, other cells, what have you) they release cytokines and chemokines. These can be either pro- or anti-inflammatory. In the case of the nucleus accumbens and drug related reward (particularly morphine, which is what we'll be talking about here), pro-inflammatory cytokines from the glia in the nucleus accumbens can increase tolerance and morphine reward.

Cytokines are coming something in to vogue in neuroscience circles. People are starting to think of them more when we think about things like stress, and when we think about issues like depression. We are starting to look at the role of inflammation and cytokines in behavior. And stress and depressive like responses are very closely tied in with concepts like reward based behavior, anhedonia, and even drug administration. So it makes sense that people looking at drug reward are going to start to consider the cytokines.

This paper set out to look at how cytokines from glia in the nucleus accumbens affect reward responses to morphine, and how both of these are affected in the presence of early life handling in rats.

Early life handling is an interesting behavior. Most kinds of early life stress have negative repercussions on rats in adulthood. But neonatal (newborn) handling is different. When you take a group of baby rats away from their mother for a short period of time, the mother gets upset, and when you put them back, she loves them up really good. Licking them, snuggling on them, nursing, all the nice mommy stuff that rats do when they are happy to see their pups. This increase in maternal care as a result of the handling means that this is not your usual early life stress. Neonatal handling is associated with reduced drug reward in adulthood. And because it's associated with reduced drug reward, the authors wanted to look at the way it altered cytokines released from the glia.

And neonatal handling had some big effects, and the differences came out when the rats were give morphine. In normally raised rats, morphine caused major increases in cytokines in the nucleus accumbens, suggested big activation of glia. In the handled rats, on the other hand, there were no major cytokine effects. They didn't really react to morphine. The clue as to WHY came out in levels of the cytokine IL-10 (interleukin-10).

IL-10 is an ANTI-inflammatory cytokine. It can reduce the effects of other pro-inflammatory cytokines. While non-handled rats show a big spike in IL-10, along with all the other cytokines, animals that had been neonatally handled did NOT show an increase with morphine, instead they had higher levels of IL-10 already, and this didn't change after morphine. In the blood of the animals, the result was similar, handled rats had blunted IL-10 responses to morphine.

All this is nice, but what we want to know is how this impacts behavior. Does this change the reward related effects of morphine? To look at this, the scientists used a test called conditioned place preference. Take a rat. Put it in a box with two separate chambers that the rat can distinguish (maybe they are different colors or feel different or smell different). Then, give it morphine and put it in one side of the box. The next day, give it saline and put it in the other side. If you keep pairing morphine with one side and saline with the other, the rat will be able to tell the difference, and when you put the rat in on the final day, it will spend its time in the side of the box associated with the feeling of reward it got from the morphine.

But not so much if you're a neonatally handled rat. Those rats didn't enjoy the morphine paired side as much, showing blunted conditioned place preference. And this was independent of the pain-killing effects of morphine or the amount of morphine in the rat's bloodstream, the neonatally handled rats were just the same in those respects.

What is going on? Why is there a higher level of IL-10 in the neonatally handled rats? The authors looked for DNA methylation. DNA methylation is the process of putting a methyl group on certain areas of the DNA in a gene. In this case, the handled rats had REDUCED DNA methylation in the nucleus accumbens, which might explain why they have more IL-10, if the methylation usually repressed the transcription of that gene (though it doesn't really make the link).

So now we have a connection between cytokine changes and morphine reward. Animals handled as pups have cytokine changes and this is correlated with a reduction in morphine reward in conditioned place preference. But this is a correlation. In order to prove that the cytokines are at least partially responsible for the changes in morphine reward, you have to prove that the presence of the cytokines is necessary or sufficient for some aspect of the behavior you're looking at. And to do that, you need to change up cytokines.

So how do you change up cytokines? Well, cytokines are inflammation associated proteins, so you give an anti-inflammatory drug. We usually think of something like aspirin, but in this case Ibudilast, an anti-inflammatory that has been shown to suppress glial cell activation before, and so was pretty ideal in this case. Ibudilast stopped the increases in cytokines that you usually see with morphine, and also INCREASED IL-10, which is similar to the effects seen in handled rats.

And how does this drug impact behavior? The scientists in this study found that ibudilast didn't stop rats from developing conditioned place preference for morphine, but it DID stop them from reinstating that preference. This is when you give a rat a few weeks off to forget the two chambers it learned. Test it a few times with no drug to make sure it really doesn't care about the sides of the chamber anymore. Then give it a test injection of morphine, and see if it still associates the original drug paired side with this new injection of drug. Normal rats without ibudilast would head STRAIGHT for the drug-paired side, but ibudilast prevented this, showing that anti-inflammatory drugs in this case can have an effect on drug reward related behaviors.

So what does it all mean? It means that morphine doesn't just act via mu-opioid receptors and dopamine (though it does that, too), it plays a role through glia in the nucleus accumbens, and that the GLIA themselves play a role in reward via release of cytokines. Not only that, it shows that the glia themselves impact the behavioral response to morphine, and play their own role in reward. If these findings hold up, it could really change how scientists studying addiction think about reward and what controls it. We are used to thinking only about the neurons in the nucleus accumbens, the dopamine, and the opioid receptors. But maybe its time we took a closer look at the glia.

Schwarz, JM, Hutchinson, MR, Bildo, SD (2011). Early-life experience decreases drug-induced reinstatement of morphine CPP in adulthood via microglial-specific epigenetic programming of anti-inflammatory IL-10 expression Journal of Neuroscience, 31