 Thank you very much, Don, it's a pleasure to be here and again I have to thank Julio for inviting me to what's been a fantastic meeting and also to take part. So I had to talk to you this morning about brain changes in schizophrenia and that includes some of the neuro pathology and some of the more recent gene expression work we've been doing. So I just want to remind you that schizophrenia is a brain disease. I think we kind of lost sight of this especially in the middle of the 20th century and that schizophrenia or psychosis is a brain disease. So I'm going to talk a little bit about the brain changes at the microscopic level. Then I'm going to introduce you to some of the work we've done on some more rather simple genomics on both the brain and the blood. How we came across a very strange protein called selenium binding protein. I think the theme of ubiquitination and ubiquitin processing is emerging throughout this talk. Then at some time I shall include some work we did on methamphetamine which was a very popular pastime in San Diego where I spent six years and was associated with psychosis. So I really think there's an urgent need to clarify that pathology underlies schizophrenia and psychosis and to me psychosis is a bit like a cough or shortness of breath or from the discussion I had this morning with Jyn and Michael and Malcolm that maybe it's a sore or inflated knee. But it's a symptom and if you went to your GP complaining of shortness of breath or a cough they'd find out what the cause is. They do various tests and tell you where you've got a chest infection, congestive cardiac failure or a tumour. We're nowhere near that level of sophistication. We're still trying to sort out what the cough is and that's a real pity and it's a limitation to our progress. If I was going to speak to you about Alzheimer's disease or Parkinson's disease, which Colin is going to have to, it would be a much more elegant talk. I'd be talking about numerous brain collections, collections large enough to have subsamples, subgroups and a really detailed advance in our understanding of the pathology including the mechanistic role of amyloid deposition, fetum gamma secretase, APP cleavage, hyperphosphorylation of tau, all these mechanisms being identified and leading to novel treatments. We're nowhere near that. So all I can offer you is this. So I think it's fair to say that at the beginning of the 21st century it was a place with schizophrenia or psychosis to say that three groups of brain changes were identified. There's a classic definition by neurochemical abnormalities affecting the dopamine pathway, glutamate and GABA. There's also very subtle but robust anatomical changes and these include subtle losses of GABAerogic interneurums. The most robust changes are the loss of markers of dendrites and synapses. One mark of dendrites is a protein called MAP2, microtubule associated 2, which has been shown to be significantly affected in the brains and schizophrenia, including in its rates of phosphorylation. I have appeared to microphotographs of nonphosphorylated MAP2 and phosphorylated MAP2 and the ratio of phosphorylated and nonphosphorylated affects the stability of the cytoskeleton. Unless you have this sort of raft of trophic signalling developmental pathways that have also been implicated including the regelin, RSG4, DLX and then a pathway we've been interested in which is the wind pathway in GSK3 beta. I'm just going to talk about this as an example. The wind pathway I think is important because it goes from birth to death. It's very important during the development of the brain. It's one of the key switches that determines whether a cell is going to become a neuron or not. Then it remains active in the idle brain in that controls the stability of the cytoskeleton, MAP2, dendrites and synaptic proteins. Both of these markers have been shown to be abnormal in schizophrenia. Then, as you age, it becomes important in neurogenital diseases such as Alzheimer's disease. The kinase which is right to the centre of this pathway, glycogen synthase kinase 3 beta, is central to mediating insulin signalling. Sometimes it takes me back to the old insulin coma therapy because insulin certainly would have activated GSK3 beta. This is the pathway where you get a wind signal interacting with a frizzled receptor. All these funny names come from mutated mice that, if the receptor was knocked out, they looked frizzled, or if the protein below knocked out, they looked dishelled. In the centre of this complex of GSK3 beta, you have a complex of proteins that, if this complex falls apart, the proteins are sent for degradation by the ubiquitin protozone system. I'll come back to the UPS system a bit later on. It's a pathway which is important to development, important to maintaining the active brain and goes awry in neurodegeneration. We actually showed in the lab a number of years ago that there was a reduction in the expression of GSK3 beta in the brains in schizophrenia. We couldn't find any changes in another protein in the pathway called beta-cutin or in disheveled. But ours and other groups have actually shown there are significant abnormalities at lots of different stages in this pathway. There's been reported an overexpression of the ligand wind, no change in its receptor or the downstream modulator protein. GSK3 beta is decreased in its staining. Also, other proteins in that complex of beta-cutin and an APC are decreased. There are accompanying alterations in the stability and expression of MAP2 as a mark of dendritic stability and the downstream synaptic proteins. This may well be one of the pathways which is awry in schizophrenia. Identifying these abnormalities may in the future identify novel targets for therapy. We were trying to work out where to go before I do that. Other groups have shown that if you decrease GSK3 beta you actually get synaptic dendritic changes. Often there is a decrease in the expression of these consistent to what we find in schizophrenia. Interesting, some of the stimulants and drugs of abuse such as amphetamine, LSD and PCP also inhibit GSK3 beta activity. Work on enriched environments and rodents has shown that this increases the ligand for GSK3 beta the wind. There is also GSK3 beta activity increases. It may be that dysregulation of this kinase, this pathway may play a part in the development of psychotic symptoms. We were trying to work out where to go with this data and we thought we would try to exploit the emerging technologies of the genomics. By that time I was working at UCSD with colleagues Ming Swong and Steve Glatt. Our discussions moved from the central idea looking at the gene expression to seeing if we could compare gene expression into compartments. We were trying to look at both mechanism and potential biomarkers. We actually put the studies together looking at the gene expression from brain tissue from the Harvard brain bank. We found some benefits made available to us. This was 19 schizophrenic patients and 27 non-psychiatric controls which were fairly well matched in terms of various variables such as gender, age, etc. Coming from Taiwan we had a sample of Taiwanese blood. We were a little concerned because ethnically they were quite different but again they were a fairly matched group. We wanted to see what would happen to gene expression in the blood and the brain in individuals with a diagnosis of schizophrenia. This is the typical way in which we actually carried out the expression obtaining the blood or the tissue extracting the RNA, producing a cDNA. Then we hybridised it to the U133 plus 2, which provides probes of the entire human genome. Out of 39,000 genes we actually found six genes with the same accession number that were dysregulated in the two compartments. I thought we were going to be overwhelmed with data and it was quite nice to actually have six probes to go after. Of those six, only one was dysregulated in the same direction. That was this gene here, selenium binding protein, and it was upregulated in both the brain and the blood. I'd never heard of selenium binding protein while I assume it binds selenium, but I'm not quite sure what else it does. We'll come back to that in a minute. I wanted to make sure that this wasn't just a tight one error. We actually validated this finding in the blood using QRT-PCR, which again showed the gene expression was increased in the brain by using immunocytochemistry. This was the typical feature in a control brain. What you're getting is a little bit, you're getting some staining in the cytoplasm of glia. You can't see the nucleus of that glial cell, but you can for this one. Possibly some faint staining in the neuronal cytoplasm. By contrast, in the schizophrenia brains, there was much more marked deposition. Again, in the glial cell cytoplasm, there's the cytoplasm, there's the nucleus of the glial cell, and the same there. Again, the possibility of some neuronal cytoplasm staining. I tend to be a bit cynical about these findings, especially as psychiatry is littered with the publications of we found this and then somebody else found the opposite. We decided if we were going to have any kind of belief or attraction in this idea, we had to repeat it again. So, fortunately, there was a researcher from Osaka, Tetsukanasawa, who was working in the lab at the time, and we asked him to work on a completely different brain series. So, this time we got access to the Stanley tissue array collection of 35 patients of schizophrenia, 35 bipolar, and 35 match non-psychiatric controls. I wanted him to see if he could actually see if there's a difference in the gene expression, selenium binding protein in that group. Of interest, 20 of the bipolar's had documented psychosis. So, we didn't bother with the affy chips at that point. We went straight to real-time PCR. Again, what he found was there's a statistically significant increase in the mRNA levels in the psychotic group by about 12%. If we broke it down by diagnostic group, the increase was 11% in the schizophrenia group, 14% in the psychotic bipolar's, and the non-psychotic bipolar's were indistinguishable from the controls. So, again, this goes to my idea of a poor to talk about doing psychosis research rather than just schizophrenia research, because psychosis where it's in schizophrenia or bipolar's order seems to blur the boundaries. So, we were quite relieved that we were actually able to validate the finding. And then when I moved to Melbourne, Melbourne has a wonderful collection of psychiatric brains, and we recently repeated the study again there. And again, we found a significant increase in the gene expression of selenium binding protein in the schizophrenia compared to the control. That's the pre-value. So, we've got three findings from three different brain series giving us the same results. So, I was also looking through the literature, and a Sophia Barnes group actually published a paper a few years ago looking at, again, selenium binding protein in various compartments, and again reported that it was increased in the gene expression in the blood. And another group have actually looked at an organic selenium compound, which they found abolished the apomorphine-induced stereotype in a mouse, which I'm not familiar with this model. They say is a model of psychosis. I'll take your word on that. But again, I thought it was just intriguing findings. Looking at SNPs, and we heard a lot yesterday about GWAS and what have you, and I can never really understand their studies, because it's a cast of thousands. There's been a couple of studies again by Canazawa looking at a Chinese population, showing that we're looking at two SNPs in the selenium binding protein that there was a significant correlation with the presence of schizophrenia. One of the SNPs was this 2800953. And then in a Japanese population, the same SNP was correlated normally with various clinical subtypes such as hepithrenic or paranoid type. And then one other final publication by an Israeli group with Bob Bellmaker has shown that there's a decrease in copy number variation in the selenium binding gene in schizophrenia. So what we're trying to do currently is get a handle on exactly what this gene or its protein does, because there's very little in the literature about it. And I'll tell you what there is in a few slides time. So we've made a plasmid of the gene. We've inserted into the GFP-tagged lentiviral vector, and we've been transfecting human primary neurons and also differentiating SY5Y cells. And we can show GFP expression, and we're trying to assess if there's any changes when we turn this gene on in terms of synaptic and enderitic markers. So this is a transfected neuron. Our big problem, and these have been a really technically challenging experience, is getting a decent antibody to the selenium binding protein. We're now collaborating with a group of prostate cancer researchers in Boston who have a huge panel of these antibodies, and so we're actually now moving forward. By interest, selenium binding protein expression is a very important prognostic indicator in prostate cancer. So we want to make sure we've got a good imaging of the selenium binding protein itself, and then this is our marker of dendrites, and we will also be marking for synapse as well. So those experiments are currently under way. The gene for the selenium binding protein is on 1Q21, which I think was mentioned yesterday as either a susceptibility loci or an area where there's a copy number variation. There's a little known about the function of this, apart from fact it obviously binds selenium, but it's not one of the classic selenium proteins. We think it's probably involved in ubiquitin proteasome system, as it interacts with a member of that system, a deubiquitin protein enzyme, and it's found in the Golgi, so we think it probably also has a role in monitoring protein folding and leading misfolded proteins to the UPS. It's also been found in the growing tips of neurites in SY5Y cells. There's been clinical links, so with the expression of the protein with cosonomas and in CNS disease, and interestingly, there's been an epidemiological link between the levels of selenium in the soil and schizophrenia, and I've been told that Australia is a selenium deficient country. So this is an analysis from 1994, and on the left here in black are those states which have the worst deficiency of selenium in either the soil or the crops grown in them. On the right are those states which are reported by Tory to have the highest rates of schizophrenia. I think with the eye of faith you could see there is some overlap. There is this state here in California, I'm not quite sure how that fits in, because there's a lot of migration to there anyway, but it's an intriguing association. So I think from hopefully the data I've been presenting to you, you believe that in psychosis there is some evidence, accumulating evidence for subtle abnormalities in the brain that affect gabrodic neurons and synapses and dendrites, and some of this may be driven by important intercellar pathways such as a winter or GSK3 beta, and that on top of that there is a number of observations now of increased gene expression for selenium binding protein which may be linked to monitoring of misbolded proteins and ubiquitination. So to complement these findings, we decided to try and change tack in terms of the way we analyzed our gene expression data, and instead of just searching for particular genes, we'd actually look for pathways. So we'd say, well we're not quite sure or we won't really care which genes are involved, but we want to see which are the dysregulated pathways in schizophrenia. And so we had by this time a different cohort of blood which we collected again from San Diego, and now with the Taiwanese series and we had a schizophrenic sample and a bipolar sample on both sites, and we used IPA to actually identify the dysregulated canonical pathways for schizophrenia and bipolar disorder. And what was very interesting was this in all four groups, the top pathway that was dysregulating was protein ubiquitination. My statistical colleagues told me the chance of that happening by chance that it occurred in all four of these was very unlikely and this was a very significant finding. As you'll see, the genes that populate these pathways are different in the different cohorts, and that I kind of expect because schizophrenia is such a heterogeneous or psychosis such a heterogeneous disorder, but I think it's more identifying the pathways that go wrong. So this brings me on to the interest that we've been leading the lab to recently which is protein ubiquitination. So obviously the role of the cell is to transcribe the DNA into RNA and then translate that into proteins, but the proteins are very complicated and they have to have a proper quaternary structure and be able to function right. And despite the help from proteins in the cells such as chaperones, many of the proteins that have folded or the polypeptides have produced are folded incorrectly. So at least a third of them need to be broken down, otherwise they will miss function and cause problems for the cell. And it's a ubiquitous pathway that occurs in all cells and you know it's not really important if it doesn't function properly in a cell that's only going to live 100 days such as in the blood, but when you've got a cell such as your nerve cell which has to exist from birth to death and last the whole time, proper functioning of this is critical. So it's the cell's vacuum cleaner and if things go wrong in the cell and misfolded proteins accumulate, you end up what we understand now to be neurodegenerative diseases. So this is the systematic view of protein ubiquitination pathway. So here is a misfolded protein and you get these E1, E2 and E3 ligases that attach ubiquitin and then transfer it onto the protein. The ubiquitin attached to the protein is a signal for it to go to this cylindrical vacuum cleaner where the ubiquitin is stripped off and the protein is broken down as a peptide that can be reused again to reform proteins. Works very well and keeps you healthy but if it doesn't work well then you start getting problems. And so we know that certain drugs of use such as Bethanphetamine, MDMA cause obstative stress. Michael Burke talked about obstative stress in schizophrenia yesterday which can then cause this pathway to misfunction and then you can start getting laying down of proteins such as alpha-synuclein, inclusion bodies, beach amyloid, et cetera. So obstative stress and a whole series of things that can stress the cell out can actually lead to misfunctioning of this ubiquitination pathway. And another intriguing lead is the involvement of this pathway in psychosis of the work most recently by Dennis Vellacoulis in my element who's interested in front of temple dementia which causes the deposition of a protein called TDP3 because the UPS system is not working properly. And he's found TDP43 to be deposited in the brain of people of FTD and psychosis as well as some people who are having a late onset psychosis. So not direct but indirect evidence that there is a possible malfunction of this ubiquitous UPS pathway that may well result in psychosis. We've more recently looked at a correlation of various ubiquitim-related genes where they're to do with activation, conjugation or degradation and their correlation with positive symptoms, schizophrenia and negative and as you can see some of them are highly correlated with very significant p-values and this one in particular also correlates with negative symptoms but there's a high degree of correlation of some of these genes with positive symptoms. Then last of all I just wanted to talk about methamphetamine. So I became interested in methamphetamine with working at the University of California, San Diego because it's very popular and the clinic I used to run had a substance abuse clinic and half the people had an alcohol problem and the other 50% was methamphetamine. People go out to desert, they cook it up, it's dirt cheap so it's very popular pastime. It also became very co-morbid of HIV. People take methamphetamine, they feel very ecstatic, very libidness and they engage in lots of risky activity but if you use too much methamphetamine it carries with a high rate of psychosis. So we looked at brains of people who died and the best character of brains to look at were people who got HIV because a lot of these people have been enrolled and studied and come to study meetings every six months or a year and we knew the lifetime amount of meth they'd used. We did a lot of different immunosidic chemical studies and finally found there was a particular nurse generation that was associated with methamphetamine in these people of HIV and intriguingly it was a lot of GABR and in this particular case it was highlighted by the presence of a calcium binding protein called calbinding which is expressed in these GABR interneurants. So this was of a control individual. This is the brown staining for calbinding. This is somebody who got HIV in kefalitis and this was an HIV infected meth user. We quantified them and there was a stepwise progression in the deterioration, the number of these and also because we had prospective clinical data we were able to correlate a lot of these neurons with worsening on memory scores and what have you. So we know they are associated with cognitive impairment and may well contribute to the fact that a lot of these people became psychotic. To complement that we also did a gene expression study on the brain tissue from these brains and found by our amazement that the genes that were massively upregulated were interferon-related genes such as ISG-15. This was certainly a quantitative, very significant increase. We could show this by staining for ISG-15 in the cells in the brains. So it certainly was a definite observation. And interestingly, ISG-15 is released in response to interferon as a tag. It identifies proteins in the cell which it doesn't recognise itself as possible viruses and tags them for breakdown by the UPS. So it's part of the UPS system. So I don't know a significant increase in interferon-related genes which are tagging viral proteins and are targeting towards the UPS. It may be part of the pathophysiology of development of psychosis and methamphetamine, but it was just an interesting observation. And as a metaphor that process, I often use this part card that's been tagged and is being led away because it's been legally parked. So we had some progress and inclusion in our understanding of the brain pathology in schizophrenia or psychosis. We're not really made that much advance over the last 100 years. This is compared with the advances in Alzheimer's disease in the last 20 years. And there's numerous leads for possible subtle changes in neuronal migration, but that is subtle, probably more affecting the interneurons than the pyramidal neurons. We have more confidence in the fact there is a disruption of cytoskeleton, especially the dendrites and the synapses, that this may well be related to disruption of various signal pathways that control those dendrites and synapses, and that on top of that there may well be an emerging disruption of a UPS function. And I think all these functions are interrelated. So I think my simple model so far, which is going to be far less simple than anything Colin shows you, is that psychosis may be related to our disruption intercellular signaling, over expression of selenium binding protein and disruption of the ubiquitin proteasim system, and they're also not yet talked about possible involvement of transcription factors, which are causing damage anatomically to silexid dendrites and gabarogic neurons, and all these are resulting in its cough called psychosis. And there's a lot of people involved in this work and I just want to thank a number of them. Thank you very much.