 Okay. Does that sound okay? Does that work? Okay. Terrific. Mark, congratulations on your exceptional leadership of NHGRI. And I want to acknowledge Colleen McBride, who has built this amazing program over the last ten years, which is really a cornerstone program that people across the country look to in terms of integrating behavioral science and genomics research. I also want to thank Barbiesekker and Colleen for inviting us to be part of this exciting event. It's a true honor, and it's a delight to see so many friends here, not only from NHGRI, but also NCI and NIDA and NIAAA. So it really shows that it's one NIH and highlights the importance of integration. I'd like to share with you our work today on nicotine dependence pharmacogenetics and personalizing treatment. And I want to start for those who are not nicotine addiction researchers with just two background slides just to try to bring everybody up to speed. So a very quick, simplified view of nicotine and the brain, so to give some context for some of the genetics research. So nicotine binds to nicotinic acetylcholine receptors, which are found on a variety of neuronal subtypes in the brain, and then this cartoon is a dopamine cell body. They bind with high affinity to a subtype referred to as the alpha-4-beta-2 subtype, but also to many other subtypes. In binding and affecting the signaling in the passage of different ions through this channel, this pore that opens up, it alters the signaling of the dopaminergic neurons, and it changes the firing of the neurons to a rapid burst firing state. In doing so, it increases dopamine release. Now the ventral, this is actually a rat brain, but the ventral tegmental area is an area rich in dopaminergic neurons, as well as nicotinic receptors. All of the cross-hatch areas are rich in nicotinic receptors, so you can see they're widely distributed throughout the brain. But importantly, these types of events occurring in the VTA, where there are important projections to the accumbens, nucleus accumbens, which is known to be an important part of the brain, part of the basal ganglia that plays a role in the anticipation and experience of reward from drugs of abuse, such as nicotine, as well as other natural rewards. And also important to some data that I'll present is projections to prefrontal cortex, which is actually bigger in us humans, but the role that dopamine plays in terms of top-down cognitive control, which is very important for behavior change. Now DSM-4 and soon DSM-5 illustrates criteria for nicotine dependence. A majority of smokers would meet the criteria for nicotine dependence, but not all smokers. And some key hallmarks of an actual diagnosis include tolerance to the effects of nicotine, meaning it may take higher doses to achieve the same level of comfort or reward from nicotine used in larger amounts than originally intended, and importantly, interfering with social recreational-occupational activities, meaning having to run outside and smoke in the middle of an important meeting because you're going through raging nicotine withdrawal. Unsuccessful quit attempts continued use despite knowledge of the health consequences, and I don't need to educate everyone today about what those health consequences are. We all know that smoking remains the greatest preventable cause of morbidity and mortality. And a hallmark that we focused on in our research are nicotine withdrawal symptoms that occur, even and particularly within the first few days following a quit attempt, including depressed mood, insomnia, anxiety, irritability, and something that we've had an increased focus on, which are cognitive deficits, and specifically, subjectively reported by smokers as difficulty concentrating or having a fuzzy brain. And we're focusing a little bit on that in terms of looking at that as a target for therapeutic efforts. Now, it's surprising, though, that despite a great deal of support going into the development through pharma and also NIH for the development of therapeutics, there are really only three classes of medications that are FDA approved for this, despite this being a significant public health problem, despite the fact that still, other than maybe in California, it's nationally about one in five individuals continue to smoke, but we have only three approved approaches. And as you can see, they help people quit smoking, but they're imperfect. So compared to placebo, and these are data that are reviewed and crossed across a number of studies and meta analyses, NRTs like nicotine patch, nicotine nasal spray, nicotine lozenge basically can double the quit rates. And you get 20% of smokers who are successful in quitting and remaining quit for six months after the target quit date. Bupropion, the next FDA approved medication, also known as Wellbutrin, which is an antidepressant treatment, works by increasing levels of dopamine and norepinephrine, has been shown to be somewhat superior to NRT, but still you're getting maybe 25%, one in four patients able to successfully quit. Varenicline or chantix, the newest kid on the block, in the news quite a bit lately for some rare neuropsychiatric events, both varenicline and bupropion have black box warnings associated with them. But even with what's considered the best in class, at least in terms of efficacy, only about 30% able to quit. So it really underscores the point that there's an urgent need to develop new therapies and improve the outcomes of existing therapies for nicotine dependence. So what I want to talk to you about today in the 45 minutes that I have, is to give you, tell you a couple of stories that illustrate how genetics research can help us get to the point of identifying new therapeutics through genetics, but also improving the outcomes of existing treatment. And I'm proud to say that we've patented a smoking mouse that, no, I'm just kidding, I wish we patented a smoking mouse. But I just really like that picture. You notice he's having his cup of coffee, which highlights the importance of triggers to smoke. But these are the two areas that I'll be covering today. So starting with how do we translate from genetic findings to new medications for nicotine dependence? So we've done a lot of genetic analysis, mainly both pathway and systems based genetic analysis. We've also done GWAS. But I wanted to highlight the results of this study in particular because it gave us some clues that is now leading to a medication development program. And what we did is we looked across 11 different genes that are important in the nicotinic receptor and nicotine system more broadly. And unlike some previous studies that focus mainly on genes that code for different nicotinic receptor subunits, because these nicotinic receptors are made up of multiple subunits. In addition, we wanted to include proteins that are important in either the biosynthesis of the endogenous ligand for the receptors acetylcholine. The receptors aren't there for nicotine, they're there for acetylcholine. And also enzymes that play a role in the metabolism of acetylcholine or the vesicular transport. So a broader kind of nicotinic and endogenous system pathway, 170 SNPs plus 329 ancestry informative markers. And this is just to show you there when you think about a system there are many different chemicals involved both in the biosynthesis of acetylcholine, the breakdown of acetylcholine, you know transporters that affect how long acetylcholine stays in the synapse and then of course the receptors that bind both acetylcholine and nicotine. And what the results of this SNP analysis pointed to was a signal on a gene called choline acetyltransferase or chat. So we looked doing both individual SNP and haplotype analysis and we found signals across this choline acetyltransferase chain. I'm going to call it the chat gene, easier to say. And it's involved in catalyzing the synthesis of this endogenous ligand acetylcholine. And we found associations of a number of different chat SNPs and haplotypes in a clinical trial with 472 European smokers associating not only with nicotine dependence but with ability to quit smoking. And I would argue that ability to quit smoking is really the ultimate clinical phenotype that we care about because this is the behavior that we want to change. Now of course these were nominally significant. I mean there were a number of different SNPs that we looked at. So we weren't really satisfied until we were able to identify, obtain independent validation of these findings. So we talked to our colleague Meng Li at UVA who had samples of European and African American smokers over 2,000 samples from a number of different families. And he was able to find also that the same SNPs showed associations with multiple measures of nicotine dependence. And some of his data is published in the first data and then a second study published as an independent replication. Now the function of these SNPs were unknown but it suggested to us the idea that in fact maybe we should be thinking about not just medications that target the receptors like varenicline but those that target levels of endogenous acetylcholine as potential effective therapeutic strategies. So how do we do that? Well there's no available chat enzyme modulator. I mean that was the gene that we focused on. But in fact there's cholinesterase inhibitor medications that are FDA approved and these serve to increase the duration of acetylcholine in the synapse allowing more binding with these nicotinic receptors that are expressed on these various neuronal subtypes. And the other feature of this that sounded interesting to us is that these cholinesterase inhibitors are FDA approved to treat Alzheimer's disease. Now I have a line of research that I don't have time to show but we have proven across multiple studies that working memory and an individual's level of top down cognitive control is an important factor in their ability to quit smoking, to ability to risk cravings to smoke and so forth. And this has been shown as being important for a variety of drugs of abuse. So we were particularly interested to find in the literature that these cholinesterase inhibitors not only improve working memory and Alzheimer's disease but also in studies of healthy volunteers. So we thought, you know, this is worth a try as a tool compound just to test whether manipulating endogenous cholin, acetylcholine helps people quit. So we looked at a drug called galantamine. It inhibits acetylcholine and it also we were interested to us because it acts as a PAM or a positive allosteric modulator at the nicotinic receptors, particularly those that have alpha five binding sites. And these are particularly sensitive to nicotine. So converging evidence sort of imperfect in terms of direct relationships but the converging evidence said, let's take a shot at it. So we don't go right into humans with this because those are big costly studies hard to fund without some good data. So we as part of having a P 50 Center of Excellence funded by the National Cancer Institute, we're able to take some money and give it through our developmental funds to a new research investigator new faculty member named Heath Schmidt at the University of Pennsylvania. And that allowed him to develop a nicotine self administration paradigm in rodents. And so you can see there's a rat with an indel dwelling catheter through which they can receive saline or nicotine. And to acquire nicotine self administration. They first do this on a fixed ratio schedule. And the animal learns to press one bar to get a nicotine infusion and another bar, they don't get anything. It's an inactive bar. And so he establishes so the rats, I guess they like nicotine, they learn to administer nicotine by pressing the bar. And what he tested is whether an IP dose of galantamine right before putting them in the operant chamber chamber versus saline would somehow alter the likelihood that they would want nicotine or anthropomorphizing but I think you get the idea. So his results he first looked at acute administration. So this is the total number of lever preface. The more they lever press for the active bar, the more they want the nicotine. So a five meg per kid dose of galantamine reduces the amount of nicotine self administration. So they're reducing their nicotine intake with no effect on the inactive lever bar. And Donapasil is also a colon esterase inhibitor. And at these two doses and somewhat of a dose response fashion, you also see the rodent self administering less nicotine in the presence of these colon esterase inhibitors. However, smoking is not a single acute event. People smell chronically. So what he did is he used a chronic administration paradigm, comparing saline versus the five or the the point five or the five milligram dose that's this dose. So this dose here actually delivered chronically showed a significant reduction over time of the extent to which the animal self administered nicotine. So this is very promising to us. He also showed in a model of nicotine reinstatement where you extinguish the behavior, and then you give them a single nicotine prime that the mice given the saline before the nicotine prime reinstated and were lever pressing like crazy after they got that nicotine dose. But if they were given galantamine, they didn't reinstate and that's considered a model of relapse also. So with that compelling evidence, our postdoc who is now a an assistant professor at Penn was able to receive a K 23 grant to study the effects of galantamine on cognition and smoking cessation in a short term study in human smokers. So she and Heath are working closely together to try to advance this. I'm hopeful that this will work. Who knows? But it's certainly a good pathway in illustrating how genetics led us to this point. Now what I want to focus on now is really the crux of the talk, which is how we use genetics research to identify biomarkers to be able to improve the outcomes of the existing therapies we have. Can we get smarter about who we give particular medications to the choice of therapies, the duration of therapies based on genetically related information. Our work in this area is based on an extremely simple premise. That is, as people nicotine a clear nicotine more quickly to and it's an active metabolites, they tend to smoke more. This is because smokers like to keep their nicotine in a very comfortable range. Too much nicotine makes people feel nauseated and not enough nicotine in a chronic smoker makes them experience withdrawal. So people titrate their smoking. And we also know that there's individual variability in the rate of nicotine clearance. And so the hypothesis is that this alters smoking behavior and smoking cessation. Now of course, it's if you want to get a complete genetic picture, you want to look at all of the nicotine metabolites. But the real action is in this red circle. Nicotine is converted to codeine and it codeine is converted to trans three hydroxy codeine. And 80% of the conversions of this metabolism of nicotine occurs right here. So we focused initially on here and subsequent research has actually shown that these other metabolites don't have significant effects. But both of these conversions, nicotine to codeine, codeine to three hydroxy are mediated by the cytochrome P 52a6. So nicotine metabolism is genetically influenced. This is work by Gary Swan and his colleagues. These are heritability studies comparing MZ and DC twins, where they were able to estimate that in terms of nicotine clearance and codeine clearance, roughly 60% of that influence is due to heritable genetic factors. Now identifying those particular genetic variants is challenging. This has worked on in Rachel Tyndale's laboratory, it's showing the frequency of the Cyp 2a6 genetic variants in European ancestry populations. These variants here, which are infrequent in the European ancestry populations, more frequent in Asian populations, are rare, but they're null variants. So if somebody has two copies of these alleles, they're not actually creating any Cyp 2a6 enzyme and will not be able to metabolize nicotine. And these more common ones are reduced activity variants. To this state, Rachel's lab has identified more than 30 different alleles. The genotyping for Cyp 2a6 is very complex, so we had the idea that there's more to it than just the genetics, and we had the idea that maybe a phenotypic biomarker would be better. In addition to the complex genetics, the reason for going to a genetically informed phenotypic biomarker is because of the importance of environmental factors that also influence nicotine metabolism rate that would affect smoking behavior. And this has worked by Neil Benowitz showing that females metabolized clear nicotine faster than males, and females on all our contraceptives metabolize nicotine even faster. So what we wanted to do is have a phenotypic marker Cyp 2a6 activity in nicotine metabolism rate, and we thought this would have greater utility in predicting quitting success and as a biomarker for targeted therapy and treatment response. So over the last 15 years through funding for the Pharmacogenomics Research Network with my colleagues Neil Benowitz and Rachel Tindale, we've set out on an odyssey to develop and validate this marker. The first five years of the PGRN grant were about some of the research that I just showed you, characterizing the Cyp 2a6 variants, assessing the analytic reliability of this marker, and in the second phase, and I'm going to show you all of this data, we did a retrospective analysis into various different clinical trials to look whether there is association between nicotine metabolism rate assessed at baseline and treatment outcome. And now I'll share with you, we're moving into the proof-of-efficacy cost-utility phase, which is sort of the third and perhaps final phase of validating a biomarker for use in clinical practice, and I'll go through this briefly. So again, the biomarker, the nicotine metabolite ratio, is based on the ratio of 3-hydroxy-containing over-containing. And it's highly correlated with genotype. These are genotype-based categories of normal intermediate and slow metabolizers by genotype, so it's correlated with genotype, but as I mentioned, it's not just about the genotype, or even this genotype, there's more going on. We wanted something easy to measure in clinical practice, and this shows that plasma measures, which were the gold standard, are highly correlated with saliva measures, so potentially you could actually have a spit test in a doctor's office to characterize someone's nicotine metabolite ratio. It's stable over time in regular smokers and independent since time last cigarette, a physician would not need to have the smoke or go smoke outside before testing it. It's a stable measure as long as they've had a cigarette in the past seven days, and importantly, it reflects both genetic and environmental effects on CYP2A6 activity. All of these features would make this biomarker easy to perform in clinical practice. So we have three clinical trial populations performed at the University of Pennsylvania. They're comparable, roughly half of participants were female, varying racial distributions with the highest proportion of non-whites in the first NRT study. On average, folks smoke about a pack of cigarettes a day and they're moderately dependent. The first question was, can we just predict with a simple patch who is gonna benefit from the patch and who will not? It was a standard clinical trial as an open label trial comparing nicotine patch for eight weeks to nicotine nasal spray plus seven sessions of behavioral group counseling, very high retention rates in this clinical trial, and here's the upshot. So for all these slides, you'll see on the Y axis is the percent of people quit. This is always biochemically confirmed. Smokers don't always tell you the truth about whether they're smoking, so that's an important point. Here, along these axes, you can see that we're separating it for illustration into groups. These are the slowest metabolizers by the NMR. A less slow, faster and faster. And this is end of treatment data and this is the six months later. So you can see here at the end of treatment with each increasing quartile of group they fall into as the NMR gets faster, as they become higher and they become faster metabolizers, the odds of quitting is reduced by about 30%, okay? And that is maintained at six month follow up even four months after they're off the patch. With the nasal spray group, we really don't see any effect because A, people can self-titrate with the nicotine nasal spray and we did find slow metabolizers gave them less, gave themselves less, fast metabolizers gave themselves more and generally it's just not as an effective treatment because it has some adverse of qualities and people don't like using it. Proof of mechanism or at least one mechanism, faster metabolizers have lower levels of plasma nicotine from the patch because they're clearing it more quickly. However, we don't think it's all about the nicotine because the nicotine levels somebody has on the patch are just not that highly predictive of their quitting success and I'll get to that. We think there's more to it than just the fact that the fast metabolizers are clearing the patch nicotine faster. Validation is important. A second study now of 568 participants and you can see the quit rates here. The slow metabolizers are kind of moving away from the pack 42% quit at the end of treatment and with these groups about half is likely to quit the second to the third quartile. Well, if you slow metabolizers do so well with just a patch treatment that they can buy over the counter what about if we extend it for six months? Why does anybody think that addiction can be treated like an acute illness when things like hypertension or diabetes require long-term therapy? So we did a trial, standard treatment for eight weeks followed by placebo patch for 16 weeks versus active patch for 16 weeks and here's what we found. In the normal metabolizers it didn't matter extended treatment didn't do anything partly because the patch is just not that effective for faster metabolizers. Slower metabolizers during the period when they're still on the extended patch which is six months, 50% of slow metabolizers if you extend their patch will be quit at six months. One and two smokers quit. That's an exceptional quit rate for people who know about getting smokers to quit. So we then wanted to say, okay, well what are we gonna be able to do for the fast metabolizers? And we also wanted to know because we had a hunch that it's not just about the nicotine is there some general liability to relapse the fast metabolizers have beyond just clearing nicotine faster? So we tested this into a bupropion placebo controlled trial, bupropion delivered for 10 weeks versus placebo and group counseling. And here's the results of that trial. We did have a significant NMR by treatment interaction. And what you can see is that this is the placebo group now. Slow versus fast nicotine metabolism is highly predictive of quitting success even on placebo. And I can say more about this in the question answer period but there's something different about the brains of slow and fast metabolizers. It could be how quickly nicotine is cleared from the brain, the rate of return to about variability of nicotinic receptors after nicotine is cleared but even on placebo with counseling only only 10% of the fastest metabolizers are quit compared to 32%. But the other thing that's remarkable about these findings, if you're a slow metabolizer why take bupropion? You get no additional benefit from this medication. You do well with counseling only or as I showed previously, counseling plus patch. If you're a fast metabolizer, don't bother with a pill that may have additional cost and side effects. In fact, in this group, the odds of success with bupropion for placebo are 5.6 times greater the odds of quitting with bupropion. So it really started to suggest to us a strategy. So to summarize, we've shown that the slow metabolizers in these groups here have a significantly higher benefit from nicotine patch than the faster metabolizers. We verified this in a completely independent study showing slow metabolizers do better with the patch. We've also shown though that these fast metabolizers you can restore their deficits, you can revive their quitting ability with a non-nicotine medication like bupropion. We also have preliminary data showing that Chantix also a non-nicotine medication has similar effects. So it leads us to the idea that a slow metabolizer would do well with patch over the counter and some counseling, but fast metabolizers really need a non-nicotine medication like bupropion or verenicline. Another point for those interested in lung cancer and the toxic effects is that it's not bad enough, not only do the fast metabolizers smoke more and have a harder time quitting, they also suck more toxins in from their cigarette. So the faster metabolizers in the fourth quartile and the smoking topography where we measure puff inhalation, inner puff intervals, they are sucking in more toxins and that's also showed here because this is NNAL, a tobacco specific carcinogen. And so they're double screwed. They smoke more, they can't quit and they smoke in a way that actually exposes them to more of the toxins that will increase their risk for tobacco related disease. Now I wanted to say a word about a cluster of SNPs on chromosome 15, the alpha-5, alpha-3, beta-4 gene cluster which has received a great deal of tension in relation to a very robust reproducible association with heaviness of smoking, accounting for about 3% of the variance in heaviness of smoking, but a clear signal that's very reproducible. So looking at that now through PNAT in a study within our team led by Andrew Bergen, he took over 2,600 treatment seeking smokers from across eight different clinical trials and combined the data for smokers receiving different treatments into single treatment arms and did this type of mega-analysis and what he found is that the risk alleles for heaviness of smoking from these two SNPs and alpha-5 and alpha-3 were associated with reduced quitting rate at six months if they were on placebo, but those same SNPs were associated with an increased quitting rate if they were on NRT and that would be any type of NRT. So in fact, it suggests that these SNPs may have some predictive validity for smoking cessation, although the data suggests that part of this is mediated through heaviness of smoking. The issue is though, there are limitations that for one thing that the grouping, it's not as clean to do this across multiple studies and pool them and also I think because of that, the amount of variance accounted for in predicting quitting is low. However, we hope to be able to address that in this study that's nearly completed. So this is our second generation pharmacogenomics research network. It's a UL1 grant. I'm pleased to say it's not as the primary funder but NHGRI is a funder of this grant as is NCI and as is NIGMS. So it's a major effort. It's an international trial for sites, I can say international because one of them's Canada. So that still counts. But it sounds more impressive than national trial. And this is the first prospective stratified pharmacogenomic trial in smoking cessation treatment, 1200 smokers where we're oversampling the slows. So 600 will be in the slow group, 600 in the normal group and then based on their slow and normals will then be randomized to placebo, nicotine patch or veranicline which is now more widely used than bupropion. So not only will this be sort of the next study to actually be on the retrospective studies to provide the evidence base for using the NMR for targeted therapy but because we'll be genotyping for all of the chromosome 15 SNPs as well as other candidate SNPs we're considering using a chip that NIH has developed to really do a comprehensive analysis. We'll be able to develop an algorithm that combines both phenotypic and genomic markers and then to test the utility of this, not only in terms of efficacy but working with our health economists to look at the cost utility of this targeted therapy and personalized medicine. I'm happy to say that we're over the 1050 subject point and we will have all subjects accrued in September with our readout of the end of treatment effects analysis by the end of the calendar year, if not sooner. So we're very excited about this. In anticipation of this, there are two companies that are funded through SBIRs that are developing point of care tests that can be done in doctors' offices quickly to be able to assign people based on the NMR. I don't have any financial interest in those companies unfortunately but we're really hopeful and we're trying to work with them and share our data so we would like to see this translated to practice. So in summary, our genetics research suggests that modulation of endogenous levels of acetylcholine may be an effective therapeutic strategy for nicotine dependence treatment and we've shown that a genetically informed biomarker of nicotine metabolism rate, the NMR, shows promise for tailoring smoking cessation treatment choices and it raises the question of then how do we translate this to practice and I've had the privilege of working with my colleague and friend, Alexandra Shields and who will be speaking next and addressing those issues. I want to make some acknowledgments. It takes a village to do these studies. First, this is the PGRN group. So here we have the little Canadian flag. I am, the best thing that, well, one of the best things that NIH ever did was create this multiple PI model so that we can share leadership and work together in that way. Our institutions like it, so Rachel Tyndale is my colleague and co-PI on the bench side and I'm on the clinical research trial side but we have a large number of collaborators. I won't take time to mention all of them. It's a really terrific team with health economists, psychologists, pharmacologists, geneticists, behavioral scientists and more. I want to acknowledge also Penn faculty who've been involved specifically in the data that I presented as well as trainees and staff who worked on the data that I shared and importantly, we could not do it without funding from friends at NIH and this is the funders of our U01 and also want to acknowledge the P50 that's been funded through NCI which has helped us get to this co-inesterase inhibitor and other medication development studies. So I thank you very much for your attention. Thank you.