DR. DUYN: Thank you, Dr. Weinberger. We have time for a couple of questions.
DR. WEINER: It is a very beautiful story you tell. You repeatedly spoke about NAA concentrations and NAA signals, but in fact your data was almost all NAA ratios, and those ratios can change, because the creatine changes, for example. So to what extent have you tried to pursue that, to demonstrate that it is, in fact, the NAA and not some other metabolite changing?
DR. WEINBERGER: One of the things we have done in all these studies is that we have looked at NAA, not just NAA /creatine, but we have looked at NAA relative to NAA in other regions. So we have used NAA in the region of interest ratioed to NAA in other regions, which has always supported the finding.
So we don't think the data are explained by creatine variation, and in general -- in general, if we ratio to choline, the data are pretty much the same. But when we ratioed NAA in other regions, it is pretty much also the same data. So we think it is the NAA signals.
We have experimented somewhat using the water signal as a reference in the water acquired datasets, but we haven't really had this extensive a dataset to look at with the water acquired signals.
Let me just make one other point. Particularly in the genetic datasets where we have three population studies, all showing virtually the same relationships, it becomes harder to associate these changes to the denominator across all these datasets; because there are different conditions. They are obviously different methodological factors, but the absolute quantification has not been done.
DR. MATALON: Yes. I have a question regarding the NAA and other metabolites that you measured and were unaffected. I am interested, really, in metabolites that you did not measure, because NAA, to me, is a marker, and what is left to be done?
DR. WEINBERGER: Well, I mean, you know, you are absolutely correct. These are crude measures of a peak which probably is hiding other signals within it, and we would love to be able to measure NAAG, and we are hoping that with some of the newer, more sensitive methods, shorter echo times and some of the editing strategies that are available now, we will be able to resolve the NAAG signal.
We are looking with short echo time acquisitions now and less imaging based strategies at a number of other signals in the proton spectra. We have not acquired those kinds of data with these strategies, but clearly, there is a lot more biochemical information in the proton spectra than just these measurements.
You know, let me just say that, from our perspective, we actually looked at several genes in the NAA biosynthetic and degradation pathway as possible measures of these signals. We don't find associations to those genes in our datasets.
We don't really believe that we are dealing with a disorder of NAA metabolism. We believe, and these studies have all been structured around the idea, that this is not a biochemical experiment. This is an experiment using a proxy measure of something about the biology of this region of the brain.
It may well be that this is an underestimation of what the biochemical signals could tell us, but that has been basically the strategy that has been used so far.
DR. ROSS: Let me just address the issue of the heterogeneity -- a beautiful presentation, very convincing -- heterogeneity in the literature regarding NAA and schizophrenia, and you refer to a poster which is actually not displayed outside. It is carefully disguised, because we don't want to disagree with you.
DR. WEINBERGER: It has happened before, trust me.
DR. ROSS: The heterogeneity actually resides in the response of Peter Barker to my question. The standard deviation of measurements in Jeff Duyn's elegant technique is between 10 and 15 percent.
In my casting an eye on the data here, and having looked at some of your published data, the difference you are looking for in the dorsal prelateral -- however you pronounce it -cortex is 12 percent max. So we are bound to find differences when we go to other techniques and other laboratories, and difficulties in anatomical placements.
So I don't believe there is a disagreement at that level. There does need to be some standardization, and we heard that also from Peter. But probably the take-home message is that schizophrenia does not give one diagnosable MRS data, unlike Alzheimer's Disease which we will hear from many speakers.
I think that is the only point here. So I agree thoroughly with your point. This is not a biochemical statement about a metabolic mechanism. This is a surrogate, and you demonstrate that very beautifully.
DR. WEINBERGER: I completely agree with you. These are -- at the level of contrasting groups, these are very weak signals. I mean, there is no question that the difference between schizophrenics and normals in these measures is very small, which is one of the reasons I think, in fact, there has been -- you know, there is a limit to the enthusiasm one can have about making the measure as a signature finding. It is, obviously, a finding that may have some predictive value in other aspects of the illness, but it is certainly not diagnostic, and it is certainly not dramatic.
DR. NAMBOODIRI: One comment regarding the mGluR3 and NAA. You note that NAAG is a ligand for the mGluR3. Do you think that there is something more to that than just mere correlation?
DR. WEINBERGER: That is one of the reasons we obviously looked at NAA as one of our target phenotypes for GRM3. I think, if we could get an NAAG peak out of these data, it would be extremely interesting to look at that.
My own personal guess is -- again given the limitations of resolution, the variance in the measurements, my own personal guess is we are measuring an integrative characteristic of probably the synaptic abundance and, therefore, oxidative phosphorylation related to synaptic activity of this region. This is my guess of why these measurements have any predictive value.
I don't think we are measuring something specific about NAA processing or about NAAG, GRM3 signaling. My own personal bias is -- and this is just based on the fact that these data work -- is that we are really measuring some integrative characteristic of the overall synaptic activity of this region, which indirectly is critically related to oxidative phosphorylation in mitochondria, because that is what is necessary for synaptic activity.
I think BDNF works, because it determines synaptic abundance, and I think GRM3 has to do with excitatory activity, and excitatory activity is critical for spine density, you know, synaptic abundance, and mitochondrial activity. So that is my guess about why these things work.
DR. MATALON: I just want to make a little comment to see what you think.
To me, I can say that, because I am not an MRI or MRS expert. So you can think this guy doesn't know what he is talking about. I don't mind that. I don't know.
To me, MRI and MRS are for compounds in high concentrations, macro concentrations. To say that we cannot see other things, a micro nutrient or micro components that may be extremely important -- we are neglecting them. I think this should be said right at the beginning, and what do you think about that?
DR. WEINBERGER: I mean, you know, if we can get -- I mean, I think we would like to be able to make more definitive measurements. I think it would be a great advance. There is a lot of interest in doing that.
You know, there are always limitations based on sensitivity and resolution. You know, we are taking a relatively large piece of brain which has functional differentiation and sticking it in a voxel, as if it is homogeneous. You know, there is a lot of error in these things, but I think it would be great if we could refine the biochemical resolution.
DR. PHILIPPART: You have mentioned the blood flow. I wonder if you could tell us more about what is known about blood flow, which is, after all, you know, a crucial way of measuring brain function.
DR. WEINBERGER: Well, again I think that is a whole other symposium, and there have been many. But our application of blood flow measurements was as another physiological assay of neuronal function. So the blood flow measurements I showed you were acquired during the specific cognitive processing of a working memory task and a sensory motor control task.
So presumably, these were activation data. So presumably, the response at the blood flow level of contrasting one state to another state and the regionality of those patterns reflects something of the engagement of the neuronal circuitry and the function in the circuitry.
So here, blood flow was presumably a measure that indirectly monitors neuronal activity. Actually, there have been studies now -- we have actually done studies with magnetoencephalography, looking at exactly the same task where we are not monitoring blood flow or BOLD response. We get the same patterns of activation.
DR. DUYN: Are there further questions for Dr. Weinberger? I thank all the speakers for their work here.
Was this article helpful?