Depression is a disease with a difficult set of symptoms. Not only are the symptoms difficult to describe (how do you really describe anhedonia, before you know the word for it?), symptoms of depression manifest in different ways for different people. One person will eat more, sleep all the time, move slowly. Another will eat almost nothing, never sleep, and be irritable and nervous. They are both depressed. The only universal symptom is the feeling of...depression, and the need for successful treatment. Treatments which often take several weeks to work, are often ineffective, and which come with a host of side effects.

So I was particularly intrigued when Nature published two papers this week looking at the role of dopamine in depressive-like behavior. What I particularly like is that these two papers have somewhat opposite results, due to different behavioral methods, something which I think highlights some of the problems associated with studying depression. Ed Yong covered both of the studies together fabulously over at Not Exactly Rocket Science, but I'd like to look at them both separately, to take a deeper look at each one, see what they've achieved, and what other questions they raise. So I will start with one today, and post the other tomorrow, looking at both sides of dopamine's potential role in depression.

Tye et al. "Dopamine neurons modulate neural encoding and expression of depression-like behaviors" Nature, 2012.

While you often hear about serotonin in studies of depression (serotonin is, after all, a target of many current antidepressants), there are many other neurotransmitters and systems that are also under investigation, and many of them are bearing some fruitful results. Ketamine, for example. And of course there is the role of dopamine.

We usually think of dopamine linked more with things like reward or drug-addiction, but what dopamine actually does is more complex than that. Dopamine is involved in movement, for example, but it is also involved in, for lack of a better word, "motivated behavior". I often think of dopamine in terms of "salience", helping to determine how relevant something is to your interests, which encompasses motivated behaviors for food, sex, drugs, etc.

And dopamine could also be important in major depressive disorder. People with depression often exhibit reduced motivation, anhedonia (a decrease in pleasure from usually enjoyed things), sometimes motor decreases as well. All of these are linked with dopamine. So targeting the dopamine system is one of the ways in which we can look at potential mechanisms and treatments for depressive behaviors.

For this study, the Deisseroth lab (famed for the development of optogenetics) used their famous approach, combined with behavioral techniques, to look at the role of dopamine in depressive-like behaviors in animals. Optogenetics is a fascinating technique. In it, you insert a gene (usually via a harmless viral vector) for a protein called halorhodopsin. This is a channel that, when activated by light, will inhibit action potentials in a neuron, effectively shutting the neuron "off" from signaling. When you insert the gene for this channel into a neuron using a virus (and usually targeted for a specific neuron set), it will be expressed, and then when you shine a light into the animal's brain, the channel will be activated. It's a fast and efficient way to shut down very specific sets of neurons.

In this case, Tye et al aimed for the ventral tegmental area (VTA), a set of dopamine neurons that projects to areas like the nucleus accumbens. This is a system that is very closely related to reward and motivation (as opposed to the other set of dopamine neurons in the substantia nigra, which is more closely related to movement). Then, when they had expression, they tested the animals for a variety of depressive-like behaviors, with the light off (so dopamine neurons were firing normally), or with the light on (activating halorhodopsin and thus turning the dopamine cells "off").

What you can see here is the result of two of the tests for depressive-like behavior, tail suspension and sucrose drinking. Tail suspension (on the left) involves hanging a mouse up by its tail, so that it cannot escape. The animal will struggle for a while, but then give up, showing what is sometimes termed passive coping behavior. If you give the animal an acute injection of an antidepressant like Prozac, the animal will increase struggling. If you expose an animal to chronic stress, on the other hand, they will show less struggling.

And apparently, this also works if you turn "off" the dopamine neurons in the ventral tegmental area. When the light was on (and the dopamine neurons "off"), the animals showed reduced time spent struggling compared to when no light was present.

Of course, because dopamine has a great deal to do with locomotor activity, they had to check that the decreased struggling wasn't just a side effect of decreased locomotor activity. So on the top right of that figure, you can see the locomotor effects when the light is off vs when it is on. While there is no statistically significant change, it does look like there may be something there (which I will get to later). But the authors also conducted another type of test. On the bottom right you can see sucrose preference, which is used to test anhedonia, or lack of pleasure. A mouse is placed in a chamber with a bottle for water and one for sucrose (which they like very much), and you count how much they lick the sucrose. You can see that when the light was on (and the dopamine neurons "off") the animals drank less sucrose, showing an anhedonia-like behavior.

Then, the authors decided to look at the dopamine neurons from another direction. In a different set of animals they did another optogenetics experiment, this time using channelrhodopsin. While halorhodopsin is a channel that turns neurons "off" when activated, channelrhodopsin does the opposite, producing increased activation of the neurons. They then exposed the animals to a paradigm known as chronic mild stress for 8-12 weeks. This is a series of mild stressors (damp cage bedding, cold environment, disco music, your cage being tilted oddly) that rotate, so the stress is unpredictable, for 8-12 weeks.

Here you can see animals exposed to chronic mild stress, and the effects of the channelrhodopsin. After chronic mild stress, animals show increased depressive-like behaviors, they spend less time struggling in the tail suspension (the left of the figure, see the grey line), and drink less sucrose (bottom right of the figure). But when the authors turned on the lights and activating the channelrhodopsin, animals that had been stressed began to look like unstressed animals, showing more struggle and sucrose drinking (see the blue line). This shows us that decreasing dopamine cell firing can produce an increase in depressive like behaviors, and increasing dopamine cell firing can help prevent depressive-like behaviors. This effect required dopamine neurons, and only dopamine neurons, influencing glutamate neurons in the same area had no effect.

So it appears that changing dopamine cell firing from the ventral tegmental area could have a big effect of depressive-like behaviors. What is particularly interesting is that these effects were immediate, onset within seconds, as opposed to the weeks required with current antidepressant treatment. So it opens up the possibility of looking at alterations in dopamine for treatment of depression in humans, something which we've already had some hints about (for example, some doctors use Ritalin to augment antidepressant treatment and make it work faster, and Ritalin has a strong effect on dopamine). And it provides a nice mechanism for producing changes in depression-related behaviors.

While this is a really interesting and good paper, making great use of coming new techniques with older ones, it also raises some interesting questions for depression related research. The results of the optogenetics were significant in producing increased activity in the forced swim tests and tail suspension test. While the locomotor effects weren't significant definitely looks like something (and the lack of effect could be due to statistics). Dopamine is very closely linked with locomotor activity...and so are the forced swim and tail suspension tests, which can both be disrupted by locomotor activity changes. So while the locomotor results may not be significant, were they small enough not to influence the test? What does the increased locomotor activity mean for the depressive-like behavior?

The same could be said for sucrose drinking, but in a different way. Dopamine signaling is linked to salience and to reward-related behaviors. And sucrose drinking is definitely a reward-related behavior. While this may actually be an advantage in this case (showing an increase in reward-related behavior may be better for depression research than otherwise), this, and the other depression tests, really make me wonder: will we ever come up with behavioral tests that can avoid all of these potential confounds? Usually scientists get around this by doing several types of test, combining (as in this case), the more locomotor-influenced tests with a more hedonia-influenced test. But, in the case of dopamine, does this really help? What does it mean? These behavioral tests are incredibly important and necessary for finding new potential treatments and the mechanisms behind behaviors, but with all of the potential pitfalls of these tests, we may never really know how well a potential treatment work to treat depression until we can test it in a human (or until we can put a mouse on a couch and understand its problems).

And of course, it also raises more mechanistic questions. What other systems besides dopamine are involves and how much so? When two patients show three constellations of symptoms, how do we know which systems might provide the best avenues of treatment?

The authors themselves put it best:

These results underscore the fact that psychiatric

diseases defined by a constellation of different classes of symptoms can

be influenced by multiple neural circuit processes. The heterogeneity

of mood disorders further complicates the pinpointing of precise circuit

dysfunctions mediating symptoms of depression. In animal

models, tests such as sucrose consumption (when depressed patients

can experience either increased or decreased appetite) and the FST

(which may involve transitions between active and passive coping

strategies) must be interpreted with caution.

But we need to find new treatments, new mechanisms, and they begin with studies like these.

Tye KM, Mirzabekov JJ, Warden MR, Ferenczi EA, Tsai HC, Finkelstein J, Kim SY, Adhikari A, Thompson KR, Andalman AS, Gunaydin LA, Witten IB, & Deisseroth K (2012). Dopamine neurons modulate neural encoding and expression of depression-related behaviour. Nature PMID: 23235822

*One of the authors also came up with a great new way to look at forced swim counts, which I did not show here, as they mostly focused on the tail suspension for the main effect. It's the sort of thing you only really get excited about if you're a behaviorist, but it's definitely got me interested. In a forced swim test, an animal is placed in a tank of water where it cannot touch the bottom, and where it cannot climb out. You record the amount of time spent swimming vs the amount of time spent floating (animals almost never sink in these experiments, they are very good swimmers). The amount of time spent swimming can be influenced by antidepressants. Usually, when someone runs a forced swim test, they videotape, either from the top of the water tank (looking down on the animal) or from the side (to see the whole animal). The recordings are then scored either automatically (automatic systems often look at velocity from above) or with a videotape and a timer (and large amounts of caffeine). This means there's some variability present in the measurements, when you're looking from the top you can't actually see the animal's feet all the time, so you can't really see how much they are swimming vs doing the bare minimum to stay up. But looking from the side, it's hard to see all the forms of motion, you can usually only see one side of the mouse. But this author took all of that out, and instead used a magnetic ring around the tank, and a small magnet on the rat's foot. When the rat's foot kicked, the magnetic ring detected it, resulting in an automated kick score. While this can't differentiate between swimming and climbing behaviors (which is actually really important for pharmacological classification of drugs), if you're looking just for changes in's pretty brilliant.