Macdonald Christie - Symposium 2003 - BMA Home -

Why is there a monkey on my back? Neural systems and addiction.

Macdonald Christie
Pain Management Research Institute
University of Sydney at Royal North Shore Hospital

I would like to thank Max for inviting me and I would also like to express my gratitude to him for creating the Brain and Mind initiative to bring neuroscientists and mental health professionals together. I hope it is a successful endeavour in the long term that won't only benefit Australia but will help Australians to benefit the whole of humanity. I trust the success of this endeavour, and I don't say that lightly.

I would also like to thank Max for scheduling me before lunch but I am actually a bit disappointed that this talk is after lunch because I would have had a larger receptive audience before lunch. If I had spoken before lunch more than half of you would be hungry so at least you would know what an interoceptive cue is. Another third of you would have interoceptive cues demanding action for other reasons and perhaps I could then persuade you that when this occurs, when we suffer from interoceptive cues that demand action, ultimately we have varying levels of control over those cues and some of us have very little or no control. It is very clearly understood now that there is a neurobiological basis to these cues and that false interoceptive cues are generated by drugs of addiction. That it is a real biological phenomenon and therefore can be attacked from a biological perspective. I would like to persuade you of that fact first, then try to dive down to a more reductionist level to persuade you that we can really make great strides in this area for a number of reasons. I believe we have already done so. For those who aren't intimately familiar with the problem, it would be good if you are hungry or needed to respond to another urgent drive because only 25% of you at most would have a personal knowledge or a personal understanding or first hand experience of what the interoceptive cues of addiction are because you suffer from the disorder. Overall more than a quarter of the population suffer from this problem. So the scale of addiction to major drugs is very large. Of course that figure is largely accounted for by addiction to nicotine which the World Health Organisation tells us will cause the most, the greatest burden of cost on health care worldwide by 2010, perhaps 2020, largely due to the uptake of nicotine addiction via smoking throughout Asia.

4% of our population have what is classified as alcohol dependence. I am not talking about using alcohol at hazardous levels but actually being dependent on alcohol. For heroin, the figure is 0.7% which doesn't sound like much but it occurs in a target population of young people in the prime of their life and produces a devastating toll. When we reached the peak of our last heroin availability epidemic then the death toll from heroin in Australia in that age group rivalled the death toll from road accidents, which is the major cause of death among young people. So it is not a trivial social phenomenon. It is not a trivial economic phenomenon, which some estimates put at over 20 billion dollars per year in this country in terms of the full health and economic cost of these dependencies.

I will try to explain some reasons for studying the neuroscience of addiction, other than that it's a big problem the reasons really come from what we have already discovered. We are now witnessing the emergence of therapeutics that can directly target the craving that individuals suffer in addictive disorders. Those developed to date have modest effect and I will go through them very briefly on the next slide.

Twenty years ago it was understood that the alpha 2 adrenoceptor had a role in heroin withdrawal. That knowledge came from cellular physiology of nerve cells in the brain, pharmacology and functional animal pharmacology. This knowledge was translated into human management of withdrawal disorders. These therapeutics have a very limited role, they are only useful during detoxification and are routinely used in human detoxification but they cannot reduced long term craving for heroin, so relapse remains a major problem with rates of the order of 80% over six to twelve months of abstinence. Not long after the discovery in the mid-1970s of the endogenous opioids that drugs like heroin and morphine mimic, it was established that those natural morphine-like systems had a role in establishment and maintenance of alcohol dependence in animal models. That knowledge has now translated into the use of Naltrexone as a treatment for alcohol dependence. In conjunction with other therapeutic measures there is no doubt in a subgroup of alcohol dependent individuals that Naltrexone has a place in reducing craving, reducing consumption and enhancing recovery from that dependency. Similar outcomes have been shown for Acamprosate, which we think targets the NMDA receptor that Pankaj Sah told us about this morning. The NMDA receptor plays a role in nerve cells in the consolidation of memory processes at the synaptic level. There is now a vast body of knowledge implicating these synaptic memory processes in addictive disorders. Acamprosate, as far as we know, interacts with and damps down the NMDA receptor under some conditions and has a similar ability to improve alcohol dependency as Naltrexone. I won't talk much yet about Bupropion but it has some role in nicotine dependency and perhaps other disorders. One of the exciting things about these developments is that some of these therapeutic drugs are starting to generalise to other compulsive disorders, for example there seems to be a role for some of these drugs in modifying compulsive gambling. It may be a minor role but we are beginning to be able to directly treat compulsive craving in a more general way.

I would like to emphasise that the therapeutic developments on the market now are just the tip of the iceberg. These treatments have come from the very beginnings of our understanding of the dependency process, largely twenty five years ago because that has generally been the sort of time lag between discovery of a mechanism and introduction of a therapy. The explosion of knowledge we have recently witnessed in this area in the past decade will no doubt lead to a better armamentarium of therapies that will help other approaches to treating dependencies and perhaps provide much better outcomes than the poor recovery rates we see at present. No more than 20% of addicted individuals remain abstinent or in control of their drug use for more than a year. The final thing I would like to say concerning the reasons for doing this, at a more social level, is that understanding the neurobiology of this process will help the cultural shift in our attitudes towards people suffering from these disorders. It was around a hundred years ago that schizophrenia was recognised as a medical disorder and it's been a battle through this century and particularly the latter half of this century, mostly in the last decade to recognise addictive disorders as a neurobiological disorder. This is not to say that it is a virtue to simply medicalise addiction but to actually put it in perspective of being a real disorder that is not just a self-inflicted or inevitable human failing or a necessary part of the human condition. It is something that we can do something about if we properly understand it and have all of the tools to deal with it.

The other great thing about the neurobiology of drug dependence, which much of what we heard about in schizophrenia research doesn't enjoy, is that it doesn't matter what mammal you take, you can make them drug dependent. We can establish very good models of drug dependency. A rat can become addicted to heroin or cocaine or nicotine or alcohol about as well as we can. I think that one of the most exciting starting points in this neurobiological understanding to come from animal models is as follows. All of the major drugs of dependence that we know, heroin, nicotine, psycho-stimulants such as cocaine and amphetamines and, indirectly, alcohol have profoundly different mechanisms of action at the cellular and molecular levels. They work in completely different ways on nerve cells and on synapses but ultimately they have a final common mechanism of action. That mechanism of action is that they all stimulate a basal neurotransmitter system, dopamine, and the aberrant stimulation of activity of that system is what underlies the way some components of dependencies emerge, particularly the components that produce relapse in humans, which is the most intractable problem. I want to show an experiment that illustrates this in humans.

The next slide is a functional magnetic resonance imaging study and it's taken, this study published in Neuron in 1997, from cocaine dependent individuals; people who routinely take cocaine and are dependent on cocaine The study looked for the brain regions that are activated after an infusion of cocaine and this shows you the behavioural rating of an individual after the infusion of cocaine. Two things are shown here. There is the picture of brain activity during the rush, that rapid feeling of intense pleasure and the high, or the more prolonged sense of well-being, satiation or satisfaction. I don't want to deal too much with the symptoms except that the rush for drugs of dependence is an important component and addicts always end up taking drugs of dependence in the ways that give them the biggest rush. There is something important about that and it may be the time locking of the act of consumption to activation of dopamine nerves. There is the response of the high and then that's followed by a low, a depression. If you take a dependent individual and put them in the same conditions and put an infusion into them but don't include any cocaine while the individual is expecting cocaine then they experience craving. What is most intriguing is that the same parts of the brain, particularly these basal regions that are involved in dopamine transmission, the same regions we have implicated in animal studies, the regions of the brain that are most strongly activated during the processes of infusion of cocaine producing the rush and the high and also activated during the craving phase. This has been demonstrated now in a number of studies merely by presenting the cues of the drug to the individual and it is very clear that the processes we find in humans parallels animal studies. What this means at a cellular, synaptic or adaptive level is not yet fully understood. There are many exciting theories about what's going on but the more or less permanent change in the functioning of dopamine nerves is no longer questioned. There is clearly remodelling of synapses in these regions; the points of communication between nerve cells. There are clearly changes in the sorts of synaptic memory processes that Pankaj Sah was talking about before lunch. There are changes in the strengthening and the weakening of synapses in these regions and there are changes in the structural morphology of the dopamine nerves themselves that are remodelling the function of regions. So we know all of these processes are going on but we are still not at a point where we can tie down those mechanisms precisely and link them in what happens to the individual addict. We do know there are long-term changes and we know there are long-term adaptations in these systems and in other systems in the brain that surely produce compulsive drug use but the relevant details remain elusive.

We can see the exact parallels between animal models and humans. We can establish all of the major components of addiction in animal models. Tolerance, which means the drug doesn't work as well when you keep taking it. Sensitisation means you get an exaggerated response in some systems when you take the drug away and reintroduce it. Dependence and withdrawal is one component that drives addiction. Abstinence craving is probably the most intractable problem because it persists for the longest time and in humans it will persist for years and in animal models it can persist for many months. It means that if you present the cues of the drug to an animal or human that their drive to reuse that drug is extremely powerful and as soon as they use the drug all of these other adaptations get re-activated or entrained much like a cellular memory. So what addiction scientists do is to try to relate these functional components of addiction to structural and functional changes in the brain.

What I want to take you through now is what our lab has done in this area and what a number of other labs in the world have done and in a couple of areas it can suggest new therapeutic approaches even though we have only just gone beyond the tip of the iceberg in our knowledge. Because these are pharmacological approaches we can test them in animals and then humans to see if they work. The approach has produced considerable success to date despite the paucity of our knowledge at the time of initial development of these therapeutics. So I want to go through is the basis of a few of the approaches that we have explored in heroin dependency.

I use the next picture because it takes me to where I started in this field. In the core midbrain of the human brain, there's a region around the fourth ventricle, the central grey and it's one of the core targets in the human brain physiologically and in the mammalian brain physiologically for one of the major drugs of abuse and that's the opiate drugs. Many years of research in our lab and many others have established that it is also one of the core regions of the brain that generates the withdrawal response; the aversive, negative response that occurs in humans when they stop taking the opiate drugs like heroin and drives them to continue using them. So we thought this would be a very good place to investigate what adaptations are taking place that underlie drug dependence, particularly dependence on drugs like morphine and heroin. This picture is an x-ray of an intractable pain patient who has electrodes surgically placed in this region. Stimulation of those electrodes can produce profound pain relief if they are in the right place and this occurs because the stimulation induces release of the natural opiates in that part of the brain. These are the natural morphine-like substances that relate to the addictive process. So experimentally, we take the drug dependent animals, prepare brain sections from them and examine the physiological properties of the cells in this target region. Then we try to narrow that down even further by saying the first place that we should look for these adaptations is the cells that are actually being targeted by the drugs and that's shown by that blue circle up here.

We look in this central grey or periaqueductal grey region in the mouse or the rat and essentially what we first did was set about to find how the drugs work on the excitability of those cells. The next diagram summarises that knowledge. We know that opiate drugs like heroin and morphine work on a G protein coupled receptor in the membrane and they shut down the entry of calcium in through voltage gated calcium channels in this region; not the same ones that Pankaj was talking about this morning. These are calcium channels that control synaptic excitability and they also open potassium ion channels. These ions flow in and out of the cell and the cell membrane shuts down and will no longer transmit information to other nerve cells. So the cells that express opioid receptors no longer respond to electrical inputs when you have taken an opioid. We established that the nerve cells involved are inhibitory cells, GABA containing cells is the case with dual recording and immunohistochemical methods that the inhibitory transmitters in this region are shut down and they are also shut down in the nerve terminals directly on the synapses and I will show you some examples of that in a few moments. Really what that does here is that releases nerves that transmit information down to the spinal cord from chronic inhibition. So opioids ultimately excite descending inhibitory processes. Opioids enhance descending inhibition of the spinal cord and that's how we get pain relief from this level of the brain, and how we get damping of a number of other functions at this level of the brain. We reached this understanding some years ago and asked the question: If these really are the cells that are the target of opioids and dependence, what sort of adaptations take place in these cells when we expose them chronically to a drug like morphine? We know that these cells express other receptors and in fact the opiate drugs can work on three kinds of G protein coupled receptors. They are very closely related in the genome but are encoded by separate genes, and they have got the distinct pharmacophores which means that specific drugs can be developed for each of them. We know these cells express kappa-receptors as well, another type of opioid receptor and some analgesics work on this receptor type but most of the analgesics and drugs of dependence we use clinically and socially work on the mu-receptor. These nerve cells also express delta-receptors and I have drawn those inside the cell in the figure because it was always a conundrum that there are lots of delta receptors in this part of the brain but when we look at this part of the brain physiologically we can never see the effects of stimulating those receptors. Something didn't match there and we now know that it's largely because these receptors are stored inside the cell. They are not expressed on the surface of the cell so when we apply a drug to the surface of the cell nothing happens to the excitability of the cell. What this shows, there is a little inset here, which is similar to the synaptic events that Pankaj Sah showed earlier. In this case we are looking at the inhibitory synaptic currents across the cell membrane so we electrically stimulate the synapses coming into the cell that we are recording from and this is the amplitude of those synaptic events. If we apply a delta-opioid receptor agonist nothing happens, if we apply a kappa receptor agonist we see really nice inhibition that we can reverse with a kappa receptor blocker. If we apply a mu-receptor agonist (the mu stands for morphine) we see very nice inhibition that we can reverse with a mu receptor blocker.

We were stimulated by some electron microscopy work by Beaudet's group in Montreal which suggested that if you actually apply mu receptor agonists chronically to cells that have both mu-receptors and delta receptors then something happens and the delta receptors come up to the surface so we put the question: Can that happen? The next slide shows a similar sort of trace except in this case we are looking at multiple events. These are quantal synaptic events and this measure really tells us what's happening in the nerve terminals impinging on the cell we are looking at. Again we didn't ever see a response to delta agonists in control tissue but after exposure to chronic morphine then we see a very nice pre-synaptic inhibition. We can shut down those synapses with an opioid drug that acts on delta receptors whereas in tissue from drug free animals we couldn't.

What we believe may be happening is that when the mu- receptor is stimulated on the cell surface it's actually translocated inside the cell and it brings the delta receptor back to the surface with it. So we set about testing that idea. I will very briefly summarise it. The way we do that is first take brain slices from animals treated with morphine for a week. The bar here shows the proportion of neurons that respond to a delta receptor agonist. In control animals nothing happens. Normally, after we give chronic morphine a very large proportion of those neurons now respond to a different type of opioid drug that didn't work in normal animals. We then postulated that the presence of mu-receptors are necessary for this process and there are other lines of evidence that suggest that's the case. To test this we then took a mu-receptor knockout animal; it doesn't have a mu-receptor gene. We shouldn't see expression of delta receptors after chronic morphine if mu-receptor-related events are necessary and we don't. To implicate trafficking in the process, one of the key steps is a linking protein that actually binds to the receptor and drags it inside the cell. The protein is beta arrestin 2. We take a knockout animal that does not make beta arrestin 2, even though the mu-receptors work very nicely, we never see the functional expression of delta receptors at the cell-surface after chronic exposure to morphine.

So what's the value of this? I don't know if the value of this will be particularly great in drug dependence. It may or may not be and that needs to be tested more thoroughly. If the cells that are showing an adaptation to chronic morphine by expressing new receptors on their surface, if they have also got a delta receptor and we know that's the case in key sires for pain modulation, then the delta receptor can be targeted by drugs. It has been shown now in other labs that delta receptor agonists become more effective after chronic morphine. Now I really don't think that's going to have enormous value in drug dependency management but it might. It might have a value in detoxification if we can damp down some of the adverse effects of the cell missing out on inhibitory actions of morphine because they are tolerant. Maybe not but certainly in chronic pain management there could be a major role for this and there are now gladly a couple of delta receptors agonists that are in clinical trials for human pain and I think they'll be useful under special circumstances

But what else happens in the morphine responsive brain cell? Something else very important happens in the cell and this is almost ancient literature in neuroscience. It was first discovered in the mid 1970s a major adaptation in the biochemistry of cells that occurs when cells are chronically inhibited, pretty much through any G-protein coupled receptor, and certainly through opioid receptors, is we see an acute inhibition of one of the messengers inside the cell that regulates protein activity and thereby regulates excitability. We can readily see this in the cells of interest. If we chronically inhibit this signalling process then the whole messenger system hypertrophies so that when you take away the agonist you get a dramatic upsurge in the concentration of this second messenger, cyclic AMP, that renders the cell hyperexcited and hyperexcitable. This biochemistry was well known twenty years ago but it hasn't been linked very well to physiological phenomena. Essentially what we've shown is that this process does occur in these neurons so we have this tolerance phenomenon but we also get hypertrophy and cyclic AMP signalling in these cells and the mu-opioid receptors are actually very good at shutting that message down as long as they are interacting with morphine but as soon as you take away the receptor activity, that is as soon as you induce drug withdrawal, then all of the biochemical adaptations in this system are unmasked and you get hyperexcitation of the neurons.

We have searched for the mechanisms responsible for this cyclic AMP mediated hyperexcitation and now know that there's one ion channel that underlies that change in the excitability in these cell and it actually happens to be a transporter molecule, GAT-1 which is a GABA transporter molecule. It is expressed on these cells because they are GABAergic. What it actually doing to the cell is making it hyperexcited and firing off millions of action potentials and producing a withdrawal response by producing aberrant behaviour in this descending output system. Now we have already had drugs in the clinic for epilepsy that act on this transporter, for example Tiagabine, but not very specifically. This drug, or hopefully more specific GAT-1 inhibitors, may actually have a role in dependency management.

The GAT-1 inhibitors may actually have a long term role in craving if the same adaptation occurs in other neurons throughout the brain which we believe it does. I did want to make a point that although we are exploring these very reductionist mechanisms in a single population of cells we believe that the mechanisms will generalise to many classes of nerve cells and there is some evidence that this is the case.

The next mechanism that we demonstrated actually had been shown throughout the brain because not only do we see adaptations in the cell body, we see adaptations in the nerve terminal. The next figure shows the sort of synaptic currents that Pinkaj has shown, but in this case they are inhibitory synaptic currents. In these experiments we take a normal mouse periaqueductal grey neuron and electrically drive inhibitory synaptic currents from synapses onto the cell, which really reflect the GABA release events. We keep the brain slice in morphine and apply an antagonist, a blocking drug, to induce immediate opioid withdrawal. what you should see is an increase in this synaptic current. But what we've done in this experiment is blocked the transduction mechanism that actually drives the inhibitory process and so you actually see no change in the in a neuron from a normal animal, and you would hypothesise that if the transduction mechanism hadn't changed after chronic morphine exactly the same thing would happen but it doesn't. We now see the unmasking of a new transduction mechanism which gives rise to enormous hyperexcitability of these nerve terminals. So not only is the cell body becoming directly hyperexcited, the nerve terminals are becoming hyerexcited directly. We have also identified the phenomenon in the miniature synaptic currents in these nerve terminals and that the actual mechanism involved is again, I won't go through the experiments proving it, hypertrophy of cyclic AMP signalling. In fact if you turn on cyclic AMP signalling the GABA nerve terminal activity is dramatically increased.

We were doing all this work in rats and we turned to do it in mice because of the availability of technology to delete or modify single genes in mice. The cell bodies developed the same adaptations as in rats but the nerve terminals seemed normal after chronic morphine, which was initially distressing but then we looked back into literature by some colleagues and they said that in the ventral tegmental area, this is the core area where these dopamine neurons arise, the same thing happened for them. Nothing happened, until they blocked a receptor on those cells, the adenosine A1 receptor. When we tested this possibility we found the same thing happening in periaqueductal grey neurons. What is actually happening in these nerve terminals is that there is so much cyclic AMP being formed in them that the cyclic AMP itself or its metabolised into adenosine which is then leaking out of the nerve terminal and inhibiting the nerve terminal indirectly through an adenosine A1 receptor. So contrary to what we saw in opioid dependent mice in the absence of adenosine antagonists, we could again see this profound increase in probability of transmitter release from GABA terminals when we blocked the adenosine A1 receptors. We've examined all the steps in this metabolic pathway, a transport pathway for cyclic AMP to adenosine in these nerve terminals and shown that if we block the mechanisms that transport adenosine or cyclic AMP out of the nerve terminal then this A1 receptor action is gone. If we block the phosphodiesterase that degrades AMP to adenosine or if we block the A1 receptor itself this damping mechanisms are gone.

So what? It is esoterically very interesting that in the region, in the place where hyperexcitability is occurring there is a little bit of adenosine out there damping down the system. In the mouse there is a lot more than in the rat and presumably in some brain regions there is more than others. But this potentially gives us a very nice therapeutic target because what we can do is not directly stimulate adenosine receptors but use drugs that are allosteric enhancers of adenosine A1 receptors. What they actually do is they don't do anything to adenosine receptors by themselves if the receptor is not occupied but as soon there is adenosine on the receptor they amplify that effect, many fold. So what we have got here is a locus specific action of a drug only where the aberrant biochemistry and physiology is taking place, where there is a lot of natural adenosine around or too much then we can enhance the action of that adenosine to damp the system down. This has worked nicely in some pre-clinical studies in animals to date and we are currently testing it in these neurons. It might also seem OK to use an adenosine A1 receptor agonist for the same purpose but when all of the A1 receptors throughout the body are stimulate the side effects are not tolerable. So, the beauty of the allosteric enhancers is that the drugs only work where the biochemistry and physiology is disturbed.

What is also very interesting about ccyclic AMP induced hyperexcitability is that this adaptation we've observed in our brain region of interest seems to be a more general phenomenon in opiate sensitive targets and does occur in ventral tegmental area, which is a key source of the dopamine neurons and in the nucleus accumbens and not only that, the same sorts of adaptations take place after chronic cocaine as after chronic morphine and in fact in this region we don't how persistent they are but they are at least persistent for several weeks in this region after the withdrawal. So they may actually have a long term role.

When I first understood the cyclic AMP literature, it was so what in terms of therapeutics? Who's ever going to develop drugs that inhibit these key second messenger systems within every cell. The side effects would be horrendous? But if there is an aberrant consequence of that, such as A1 receptor stimulation at specific brain sites, it is targetable by specific pharmacologies. I hope that's what I have demonstrated today with both the adenosine receptor and the delta receptor, where we have some novel therapeutic potential. We need to test these possibilities and there are many other areas understood in adaptations in addiction that hold great therapeutic promise. This has really just been a few vignettes of a couple of possible targets. There are many other areas that can potentially be exploited and may prove beneficial to some humans. I will now just thank the people in my research group who have done all this work. Elaine Bagley for the postsynaptic work, Billy Chieng did the first work in rats, Mark Connor the single cell work, Steve Hack established that adenosine was key in mice, Susan Ingram showed the transduction mechanism and presynaptic hyperexcitation in rats together with Chris Vaughan.

Thank you.

Are there any very quick questions for Mac as we do have four talks to get through this session?

Q: - is addiction a chemical genetic or learning process?

A:

It's both. Let's start with the genetics, there's no doubt that there is a genetic load. I've got a genetic load to alcohol and nicotine dependence because I'm a fast metaboliser of both drugs and it means having tried both drugs I don't suffer many adverse effects because I wash them out of my body really quickly so I just get the pleasurable effects. That's established in the literature and those metabolic predispositions are very well understood. There are certainly nervous system genetic predispositions to induction and maintenance of dependencies that are not properly understood. Just like we heard with schizophrenia this morning, women's lod scores jump out in some studies and drop back in others but there is something there because if you look at twin studies, there is clearly a strong genetic load. Is it a learnt behaviour? I didn't really go into this but these basal forebrain systems are all about learning behaviour. In fact what I believe actually happens if dopamine is activated in these systems is that the ongoing behaviour at the time that dopamine is activated is reinforced and it's clear that there is a long term sensitization of those synapses that are going through the cortex in that system and that is where I believe the learning takes place and that's where both behaviours get strengthened. People move into the most aberrant patterns of behaviour because those systems are being hijacked. So is it learnt? Yes it is learnt but it is learnt at chemical level primarily.

© 2004, Brain and Mind Australia Inc. - Copyright Notice -