SciCafe: Just Can't Get Enough - Addiction & the Brain

SciCafe: Just Can't Get Enough - Addiction & the Brain - Transcript

Edmund Griffin (Assistant Professor of Clinical Psychiatry, Columbia University):

Good evening. I'm very excited to share with you today some recent findings from my research on the molecular basis of drug addiction vulnerability. Let's start with the question at hand. Why is it that some people just can't get enough? As a clinical psychiatrist I do get this question a lot. And I think that the very wording of this question actually captures some of the frustration and also, to an extent, some of the stigma that's associated with substance use disorders. But the question also captures one of the most confusing and frequently misunderstood aspects of addiction, which is this; when it comes to substances: alcohol, cocaine, exactly how much is enough? Where do we draw that line in the sand? For alcohol, is it two drinks a night, three drinks a night, four? Or for cocaine, is it two lines of cocaine a night, four lines of cocaine a night? Or does any type of routine cocaine use mean that you are addicted to cocaine? That you just can't get enough.

How much is enough? Now, to be fair, as a psychiatrist, I can tell you that this problem with precision, it's not something that's unique to the diagnosis and treatment of addiction disorders. Sometimes, I will get questions about this with regard to anxiety or depression. A person might say, "Dr. Griffin, I can't stop worrying about this mole on my arm. Do you think I have an anxiety disorder?" A little bit of anxiety and worry about the mole is a good thing. But how much anxiety is enough? How much is too much? Or, similarly, a person might ask me, "Doc, I've been sad for months since my company failed. Sure, a little bit of sadness and self-reflection might be good. It's my fourth company to fail. But, wait, Doc, how much depression? How much sadness is enough?" How much is too much?

Often, in the diagnosis of psychiatric disorders - this is including addiction - the answers to these questions lies not in the quantification, but rather in the function. How are these factors affecting your life? And which direction are they moving the needle with regard to your relationships, with regard to your work function or your time at school or to your physical health? So, when we start to think of substance abuse from the functional perspective, things start looking differently. In medicine - one of the most important lines in the sand to distinguish addictive drug use from recreational drug use is compulsivity; continued drug use despite negative consequence to your physical health, to your work life, or to your relationships. And when we look at addiction from this view, as opposed to the arbitrary definition of enough, something very interesting emerges.

We find that only a small percentage of people who use a drug of abuse will transition into addiction. For alcohol, that percentage is 12 percent. For marijuana, it's 6 percent. And for cocaine, it's 15 percent. Think about that for a second.

All right, when you consider the social and economic cost that addiction takes on our society, these numbers are strikingly low. What it tells us is that if we could make even a small dent into these percentages we could be looking at a very, very important impact in our society. So, what is it? What are the factors that make one person resilient to cocaine and another person vulnerable to addiction?

Well, the first step to answering these questions is to look at the risk factors. And some of these are listed here on the left: history of trauma, early onset of use, poverty, history of depression, personality disorder, early use of alcohol, marijuana or nicotine, marital status, race and ethnicity. So, if you look for cocaine addiction, and you Google any one of these terms, you'll find a study that looked at some connection between the two. But when you do the meta-analysis or you look to see what is the most strongly associated factor, what factors are most reproducible in the epidemiological studies, the two factors that stand out most strongly associated with subsequent cocaine addiction are prior use of alcohol, prior use of nicotine.

This observation was actually first made as far back as 1975. It's often referred to as the Gateway Hypothesis. The Gateway Hypothesis is a population-based observation that describes the sequential pattern in drug abuse progression. Use of alcohol and nicotine usually precedes the use of marijuana, which typically precedes use of illicit drugs like cocaine. Most of the public policy debate, as you guys are all aware, will tend to focus on marijuana as the quintessential gateway drug. But the epidemiology is clear that any one of these three drugs, either individually or in sequence, is a very strong predictor of subsequent drug use and abuse.

But we have to tread lightly here because this is an epidemiological study. It's a correlation. As all of you know, correlation does not equal causality. In fact, in the 40 years since the Gateway Hypothesis was first described, we still didn't know whether or not the order of the different drug classes is socially determined or if it's biologically determined. For instance, it could be that smoking marijuana and drinking a lot, you're more likely to meet somebody who happens to have cocaine in their pocket. That's not biological. But if it is biological, what are the cellular and molecular mechanisms that might mediate this process? Addressing these two questions could reveal new insights into the developmental trajectory of addiction. And it could also help us identify new molecular targets for treatment.

Fortunately, research or breakthroughs - actually - in the biology or neurobiology of memory formation have provided an ideal platform for us to begin to address these questions on a biological level. What does memory have to do with addiction? Accruing evidence over the past 20 years has actually pointed to the idea that addiction is a disease of learning and memory. This view stems from the observation that both the neuroanatomical pathways, as well as the fine-toothed molecular pathways that mediate formation for a natural reward, such as food, water or mating, these same pathways are exploited by drugs of abuse. So, let's look at the neuroanatomy first.

This here is a cross section of a human brain. In the context of a natural reward - it's a nice big piece of turkey or a new love interest - dopamine is released to different regions of the brain that will then work in concert to help you remember both the environmental queues, as well as the sequences of behavior that helped you obtain that reward in the first place. Dopamine that goes to the nucleus accumbens helps to consolidate memory to increase your motivation to perform certain tasks in the context of environmental queues. Dopamine going to the dorsal striatum increases habit formation. This is so that you don't have to learn how to use chopsticks every time you eat sushi. Dopamine going to pre-frontal cortex provides executive control over both motivation and habit formation. This makes it so you don't start eating your sushi in front of a big tiger. This is a choice mechanism, and this is what's used in the formation of natural reward. This is very important. That dopamine here is not the pleasure chemical that you might read about in the news. Dopamine isn't there to make you feel happy. Dopamine's role is to act as a reinforcing agent to make sure that you do it again. It's critical for evolution that you eat, that you drink water, and that you make babies. And dopamine is there to make sure that happens.

What we know and what we've learned over the past three decades that all drugs of abuse: alcohol, nicotine, cocaine, including powdered sugar, all of those act by increasing dopamine in each of these regions. The trick here is that whereas dopamine levels will start going down over time with your new love interest, they don't go down with cocaine. Cocaine hits the spot each time, essentially tricking the brain into thinking that it's found new love. Now, most of what we know about addiction centers on the nucleus accumbens, and that's what we'll focus most of our talk today. But if you take the nucleus accumben cells, take it out, crush it up and look at it on a plate, you find that some of the fine-tooth molecular pathways that are used to form memory - for a natural reward - these are also exploited by cocaine or addictive drugs.

So, your brain - your neurons - adapt and respond to the environment in a very unique way. Whereas, your muscles can get bigger when you exercise or get smaller if you don't exercise, your brain doesn't have that option. Because it exist within the confines of your skull, it doesn't have the option of hypertrophy. So, your brain - your neurons - adapt to the environment by growing synapses, the more synapses, the stronger the memory connection. The formation of memory in a cell precedes in two steps. These two steps are, one, a signal transduction pathway that sends a signal into the nucleus. You can think of a signal transduction pathway as molecule A hits molecule B hits molecule C, kind of like a string of dominoes.

But the point is to communicate something that happens in the environment to the nucleus, which is the control center of the neuron. Once that information is communicated, it activates a molecule here called CREB, which binds to the DNA, which is the template that holds all of your genes, and drives new gene expression, new proteins, that then create a stronger synapse. This is the basic two-step formation of memory. Well, cocaine has learned this two-step formation process. And they use exactly these two steps in the brain when you're exposed to drugs.

When you receive cocaine, it initiates a signal transduction pathway that sends the signal to your nucleus. This creates new gene products,and those gene products go on to alter synaptic connections. We now know several of the molecules that are involved in this cascade that are critical to the formation of addiction.

How do we use this information to understand the Gateway Hypothesis with regard to addiction? One way to do this would be to translate the epidemiological observation into a rodent-based study where we can do things in the laboratory, where we can actually take their brains out. This would have several advantages. Number one, it would allow us to address the Gateway Hypothesis in the absence of social factors that are already known to influence the development of addiction such as availability or your peer group. So, in this case, they're rodents. They're in a cage, and they're doing drugs. So, we can test the biology a little bit more specifically. Number two, because they're rodents, we can take their brains out. Can't do that with people. This allows us to test the cellular and molecular mechanism of addiction to understand how sequential exposure to different drugs might affect things on the biological level. And, finally, and most importantly, we can do causality studies. We can use pharmacological or genetic interventions where we change the animal's genes. We can use these types of interventions to test specific molecular hypothesis with regard to how addiction is working. And this is something that you cannot do with epidemiological studies or with humans in the lab.

So, how do we create an animal model? Not as easy as it sounds. As many of you already know, we've seen a revolution in what we know with neuroscience and how the brain works. But translating that revolution into actual clinical interventions in psychiatry has been pretty slow. One of the reasons for the discrepancy in progress is the absence of adequate rodent models. We can model something like stroke in a rodent. Just clamp the artery. You can model something like asthma in a rodent. But how do you model depression in a mouse? Or what does a paranoid delusion look like in a rat? It's very difficult.

In fact, the same difficulty exist in the study of addiction. For the past 30 years, the gold standard model for studying addiction in rodents has been rodent self-administration. So, this is a picture of it. This is a rat. And he presses a lever. It looks like he's snorting from the lever, but he's actually going to press it. When he presses that lever, it activates a computer, which sends a signal to a pump. And that pump will inject a quick dose of cocaine directly into his bloodstream. So, in this model, the animal chooses when and how much cocaine he uses. And they will do the cocaine. They'll work for it, which is great. Looks like addiction. No, it doesn't. That's not addiction.

As we discussed earlier, simply using cocaine, simply liking cocaine is a big difference from being addicted to cocaine. As we said earlier, 85 percent of people who use cocaine will not become addicted. And that's not a recommendation for anybody. But those are the numbers. Eighty-five percent will not become addicted. So, in order to study addiction, we need to improve the model.

So, when I was a resident in psychiatry I started working in Eric Kandel's lab, and I started trying to develop a model to see if we can get the animal to press the lever for cocaine even under aversive consequences. To do that, what we did is we inserted a mild foot shock of increasing intensity every 40 minutes during his cocaine self-administration session. During the first 40 minute session there's no consequence. He presses the lever one time, two times, three and four. Now, on the fifth lever press he gets an injection of cocaine. One, two, three, four, five, cocaine, and then he kind of runs around in a circle, looks high for a while and he'll pause. And about three minutes later, one, two, three, four, five, then he'll run around again in a circle. He'll do this for 40 minutes. Just a regular day of cocaine. Everything's fine.

After a time-out, the next 40 minute session is different. He presses the lever and a blue warning light comes on. He looks at the warning light; means nothing. But he keeps pressing, but on the fourth lever press he gets a mild foot shock of 0.1 milliamps. It's not an intense foot shock, just to let you know. It's your finger in a lightbulb. Even less than that. He gets the mild foot shock. He'll slow down, but if he continues pressing he'll get his cocaine injection. What this does is it forces the animal into a conflict. He wants to use the cocaine, but the cocaine comes with a consequence. And our hypothesis is that prior exposure to a gateway drug like nicotine or alcohol will make some animals more compulsive in this phenotype. To test that hypothesis, we use the following paradigm: we have the animals receiving either alcohol every day or a water control.

After 21 days of these two treatments, we introduce cocaine self-administration, which happens concurrently with their alcohol treatment. Now, importantly, the alcohol is given for only two hours a day. We want the alcohol to be fully metabolized by the time that they do cocaine self-administration. And that's important because we don't want the two drugs to see each other, and also because if they're drunk they get in there and they don't even press the levers. It's a very difficult study. So, they do not have alcohol in their system when they're doing cocaine. And we'll get back to that a little bit later.

So, does prior exposure to chronic alcohol increase compulsive cocaine self-administration? The answer is yes. If you look at the water group - these are our alcohol-naive group - even with the small challenge of 0.1 milliamps they decrease their lever pressing for cocaine by 85 percent. That's a huge drop off. And then as you increase the severity of the punishment, or the consequence, to 0.2 milliamps the majority of animals start dropping out of the study. They feel that shock once, done. By the time you get to 0.3 milliamps, they've stopped.

In contrast, animals who have been previously exposed to alcohol stay in the study, pressing five times more with the mild foot shock of 0.1 milliamps. And as you increase the intensity to 0.2 milliamps, they're still pressing it 20 percent. And then when you get to 0.3 milliamps, they're still in the study. This substantiates our hypothesis that prior exposure to alcohol does increase compulsive cocaine self-administration.

But, wait a minute, alcohol's a potent neurotoxin, as many of you might know. It kills neurons. It could be that being exposed to alcohol made it so that the animals couldn't even feel the shock on their paws. Or maybe the alcohol had some other type of cognitive effects, so that they can't interpret the blue warning light. To test for these types of non-specific effects, we use the same paradigm - one group receiving alcohol daily and the other group receiving water- but instead of having them press for cocaine, we asked them to press for a sugar pellet; a natural reward. And we asked does prior exposure to alcohol enhance compulsive pressing for a sugar pellet? And it did not. Alcohol treatment did not enhance compulsive lever pressing for a natural reward in food restricted animals. As you can see here, as you increase the intensity of the foot shock from 0.1 to 0.2 and 0.3, they decrease their lever pressing by 50 percentat the 0.2 milliamps. And by the time they get to 0.3 they're down to about 10 percent. But, importantly, there's no difference between the alcohol pretreated and the water control.

What this tells us is that alcohol is acting as a gateway drug, and that it does not have non-specific effects with regard to natural reward. There's another take-home point here. If you look at the gray line, just look at the water control. These guys have never been treated with alcohol. These are just regular, every-day rats. And see here that they stay in this - for natural reward, when they're food restricted they're actually pretty compulsive. They'll stay in the study and tolerate a foot shock way more than they do even for cocaine. And this is important because it gets back to what we were talking about earlier that evolution has equipped the mammalian brain with the machinery to navigate this type of conflict. The machinery can navigate a conflict between wanting something important, like food when you're hungry, and the consequences. So, if I told you about my friend Joe who lost three fingers and sustained third-degree burns on his body while going through a fire to save his daughter, you'd say Joe's a hero. He did the right thing.

But if I told you that Joe lost three fingers and sustained third-degree burns while saving his cocaine from a big house fire, you'd say Joe had a problem. Joe has a problem with cocaine. And what we see here is that like the human condition, simply using cocaine does not put you in that category. Whereas, the same rats have a capacity to make this decision when they're hungry. What tips them over here is prior exposure to the gateway drug alcohol.

So, what we have here is a full developmental trajectory from a risk factor - alcohol use - all the way to cocaine addiction. This puts us in a very unique situation where we can now ask what's going on in Joe's brain. What's happening there to tip him over towards addiction? Well, we can begin by looking inside the nucleus accumbens, which, again, we have a large baseline of information from previous studies.

Now, as I mentioned before, cocaine exposure alters gene expression in the nucleus accumbens. Now, for several decades we thought this was the end of the story. If you have the genes for cocaine and you use cocaine, that's a wrap. But we now know that DNA is not destiny. That there is many other aspects controlling your genes. So, you might have the genes to be tall, but that doesn't mean you're going to be tall. You can have the genes for diabetes. Doesn't mean you're going to get diabetes. You can have the genes to live to be 100. Doesn't mean you're going to live to be 100. There are other layers that control our genes. And what we found over the past very recent, maybe decade or so is that all of your genes exist in an inactive or active state, depending on these epigenetic modulators. Epigenetic is above the genes. It's another layer of control. So, your DNA isn't just sitting there waiting to be activated by the environment. There's several layers that control or decide whether or not your genes will exist in a silent form or in an active form. But two big controllers that we'll talk about today are the HATs and the HDAC.

The HATs are called histone acetyltransferases, and what they do is they add these negative charges to the histones. Your DNA is tightly wound around these balls called histones. Adding a negative charge causes the DNA to repel against itself because DNA has a negatively charged backbone. So, histone acetyltransferases cause the repelling and put the DNA in an unwound or permissive state. Histone deacetylases do the exact opposite. They rank these negative charges off, causing the DNA to go back into a coiled inactive form. Every cell in your body exist in some dynamic process between these two classes of molecules. Amir Levine a post-doc in Eric Kandel's lab, used this information to ask a question about nicotine, which is one of the other gateway drugs that is strongly correlated with subsequent cocaine use. And he found that nicotine inhibits histone deacetylate activity, which essentially releases the brakes, causing the DNA to become increasingly permissive, which makes it more acceptable for cocaine; basically opening the gates, so to speak, for cocaine-induced gene expression.

So, using this information we're intrigued by the possibility that alcohol might have a similar mechanism. After all, epidemiologically speaking it is a very strong gateway drug as well. So, we wanted to know does chronic alcohol exposure enhance cocaine-induced gene expression. To test this hypothesis we gave the animals 10 daily exposures to alcohol. And the other group received just water control. After 10 days of alcohol exposure, we gave them a one-time injection of cocaine. And then we asked does that prior exposure enhance cocaine-induced gene expression? And if it does so, does it do this by increasing histone acetylation by decreasing HDAC activity? Let's go through each one in turn.

We found that chronic alcohol use does indeed enhance cocaine-induced gene expression. And what we're looking at here is DeltaFosB. This is a protein that's already previously been shown to be one of the molecules that goes on and is involved in later synaptic transformation.

What we found, though, is that animals who've been previously treated with alcohol have enhanced cocaine-induced gene expression almost twofold above our alcohol-naive group. So, our next step is to start working backwards. Okay, if it's increasing cocaine-induced gene expression, how?

So, first we asked is it doing this by adding these negatively charged molecules to the histones? So, we did an assay to measure these histone acetylation aspects, and we found that prior exposure to alcohol does indeed increase the production of these negatively charged molecules on the histones around the DeltaFosB gene that we were just talking about. Then we asked does it do this by inhibiting histone deacetylase? And, yes, the answer was there true also. Chronic alcohol use decreases histone deacetylase activity in the nucleus accumbens.

But which HDAC? I initially had HDAC as one little ball. It turns out there are 11 of them. If we could figure out which histone deacetylase is inhibited, this would be a very important step towards understanding what could be a specific target for treatment. So, we actually went through several steps to determine which one of the HDACs is being involved. We actually focused a lot of our attention on HDAC4 and 5. We found that chronic alcohol use is acting by decreasing nuclear accumulation of histone deacetylase 4. So, again, we used a 10-day exposure protocol, and we took the animals' brains out at different time points after their last sip of alcohol. As we talked about earlier, exposing animals to alcohol causes a decrease in histone deacetylase activity. And as you can see here, as alcohol's leaving the blood stream there is a precipitous decline in overall histone deacetylase activity. If you take these same lysates - these same samples - and run them on a gel and look at the actual amount of protein for histone deacetylase 4, you see that HDAC4 is declining precipitously during the withdrawal phase.

So, this tells us that we have a specific target that alcohol is acting on to release the breaks towards cocaine addiction. This is very important. It says that HDAC4 could be acting as a node to increase vulnerability to cocaine addiction. You may have heard the term that your genes load the gun, but your lifestyle pulls the trigger. This has been a statement that we've heard in all forms of medicine. Not just in brain science. Liver, lung, kidney. This is us in medicine moving away from this idea of genetic determinism and starting to accept and acknowledge the fact that your genes are being regulated on so many different levels, and that these levels are exquisitely sensitive to the environment.

What we're proposing here is that the lifestyle choices chronic alcohol use, chronic nicotine use, and that the trigger is HDAC4. But to test that hypothesis, we would need to inhibit histone deacetylase 4 independently. We want to pull the trigger independently of that prior exposure. So, one way to do that is to use a molecule called MC1568. I was very happy when we found this molecule. It was actually produced by chemists with specific intention of decreasing HDAC activity - HDAC4 specifically. It's a selective histone deacetylase inhibitor. So, we said if we give the animals this HDAC inhibitor, if our theory's correct, giving them this molecule - no alcohol, no coke, no nicotine - it should recapitulate the cocaine addiction phenotype. And that's exactly what we found. That selective degradation of HDAC4 enhances compulsive cocaine self-administration.

This first slide here is just looking at natural reward. Again, we want to check to see if this drug is causing non-specific effects. And we found that MC1568 does not affect compulsive lever pressing for a sugar pellet. The animals at 0.1 milliamps, there's no effect. And by the time you get to 0.2, they decrease by 50 percent down to 20 percent. But there's no difference between these groups. But when they're lever pressing for a cocaine reward, you see a split. At 0.1 milliamps, there's no difference. They've decreased their lever pressing by about 15 percent. So, they're both sensitive to the shock. But as you get to a more intense shock, the two groups start to divide. Our control group has decreased down to 40 percent. Whereas, our other group is still at 75 to 80 percent. And as you get to a lower intense foot shock, our control group starts to drop out of the study. Whereas, our guys who've had selective inhibition of HDAC4, they're staying in even at a very intense foot shock of 0.4. So, this completes the loop. This is the causality experiment that I was describing earlier.

So, in conclusion, we found that animals with a history of chronic alcohol use have enhanced compulsive self-administration, which is used despite negative consequence, which is the line in the sand that we use to distinguish recreational drug use versus addictive drug use. We found that alcohol, like nicotine, inhibits histone deacetylase activity in the nucleus accumbens - the region that controls motivation. And this results in a permissive environment for the genes, which allows for more or enhanced cocaine induced gene expression. We found that chronic alcohol use promotes this decreased nuclear accumulation of HDAC4. And, finally, we show that selective degradation of HDAC4 with the inhibitor MC1568 recapitulates the compulsive phenotype.

Thank you very much.