Thursday, May 22, 2008
The Government’s Response, continued
Direct Examination of Dr. Dean Jones
[Ms. Renzi] Dr. Jones is a professor at Emory University. Dr. Jones has a Ph.D. in biochemistry and worked in the department of biochemistry for about 24 years, and he now works in the department of medicine. He had a fellowship from the Nobel Committee on research in medical toxicology. He is a peer-reviewer for a number of journals; he has been in NIH study sections. He has a number of grants, including one on oxidative stress. He directs two laboratories, one on biomarkers for researchers and his own research laboratory. He has published more than 325 peer-reviewed articles and chapters. He’s published over 100 papers on oxidative stress. He lectures widely and internationally.
d.Q: Have you ever testified in a legal proceeding before today?
Q: Are you an expert in mercury or heavy metals toxicology?
Q: But are you an expert in oxidative stress and sulfur metabolism?
Q: Did you listen to Dr. Deth’s testimony and read his report?
Q: What is sulfur metabolism?
A: [Description of sulfur amino acids; it’s ubiquitous in living systems] The most abundant thiol is glutathione.
Q: Can you explain to us as simply as possible what sulfur metabolism is?
A: Sulfur is the fifth most abundant element in biological systems. Pretty much all of life depends upon sulfur. Proteins, which are 20% of our body, all contain two sulfur-containing amino acids: methionine and cysteine. Sulfur is ubiquitous in living systems and more or less an essential component.
Q: We’ve heard about glutathione and its role in sulfur metabolism and detoxification. But glutathione does more the detoxify heavy metals, is that correct?
A: Yes. Glutathione has a very important role in metabolism. It is the major thiol and major sulfur-containing element. I haven’t gone over the various forms of sulfur, and I can do that if you like. Sulfur metabolism can get extremely complicated. What we’re most interested in is the thiol form, reduced form of the sulfur that would constitute 1/3 of the sulfur. The major thiol form is glutathione in the body.
Q: What is the role of glutathione in the body?
A: This brain scan image taken with MRI is used to measure the different chemicals in the body. Here are the major detoxification functions of glutathione. One that received the most attention over the past 50 years is its function as an anti-carcinogen. Glutathione is the counter to chemicals that would otherwise cause mutations in the DNA and cause cancer. Fifty years ago it was recognized that many different chemicals that we are exposed to are activated in the body. The most central way that the body gets rid of these is by glutathione reacting to them.
The second function is as an antioxidant, discovered in 1957, including for the elimination of peroxides. Hydrogen peroxide is produced by the body all the time. About 1% of all the oxygen we breathe is converted to hydrogen peroxide. A large portion of that is eliminated by glutathione. It does it more or less silently because we have so much glutathione and its function is so efficient.
The third function is the coenzymatic function of glutathione. Glutathione is involved in several other aspects of metabolism. [detoxification of formaldehyde] As long as you have glutathione, it will bind to formaldehyde….
The main point is that glutathione has many functions, and these are competing functions, but the way these mechanisms have evolved is to allow them to work despite fluctuation in glutathione content. There are natural variations and large number of reactions, and in order for them to work, the system has to be designed so that activation of one doesn’t inactivate the other. Glutathione is able to support all of these different functions.
Q: And glutathione is found in every cell of the body?
A: Yes, we synthesize it in every cell. It’s composed of 3 amino acids: glutamate, cysteine and glycene. The capacity to synthesize is very high in cells. If you look at different organ systems, 10 millimolar is what you would see in a few tissues--liver, kidneys—but many cells, like red blood cells, only have .2 millimolar, and in the small intestine if in a fasting state, it’s only .1 millimolar. So there are different concentrations in different systems. We have so much glutathione that even at that lowest level, we have plenty to carry on these functions.
Q: Dr. Deth’s hypothesis is that thimerosal containing vaccines (TCVs) disrupt sulfur metabolism, and these disruptions cause autism. Is that your understanding of his hypothesis?
Q: When you analyzed it, you asked three questions?
A: Yes. There are different aspects to this, at least from my expertise in sulfur metabolism. The first is whether thimerosal in the doses administered would have significant effect on sulfur metabolism.
Q: And based on your knowledge and expertise, do you have an opinion that thimerosal in the doses it’s administered in, can significantly affect sulfur metabolism?
A: I do have an opinion. The data show that the dose of thimerosal will not affect the sulfur metabolism.
Q: Why is that?
A: …The total body thiol is taken from older literature where scientists took cadavers and measured in different organ systems how much glutathione is there. They estimated in various ways, about 800-1,000 micromolar, which is pretty consistent in the literature. Compared to nutritional recommendations of 0-6 month old child, that would be equivalent to 500 micromolar per kg of body weight. What that tell us is that on a daily basis, we take in a lot of sulfur. That sulfur is equivalent to about half of the total amount of glutathione in the body and relatively a much smaller fraction of the thiol in the body. We have a huge thiol content in the body; glutathione is an important component of that.
Q: Body thiols also detoxify heavy metals?
A: That’s correct. All thiols are potential bonding sites for heavy metals…. The RDA is set at 500, and then you have to ask how much do people actually take in? The RDA is set to guarantee that nobody has any deficiency, so it’s higher than anyone would need. The range of values is 1%ile to 99%ile or 250-500 micromoles per kg of body weight.
Q: Special Master Hastings: What does RDA stand for?
A: Recommended Daily Allowance.
Q: Special Master: Is that a maximum or a minimum?
A: When they set this, they set this value so that they expect no one to have deficiency at this point.
Q: So if you get at least 500, you’re not going to be deficient?
A: That’s right. And so in that sense, it’s more than we would need. So some individuals may get at the lower end and may benefit from a little more, but there is essentially no sulfur-amino acid deficiency. As long as child is fed, they’ll get sufficient sulfur amino acids.
Q: What are the sources of sulfur amino acids in diet?
A: Animal products are particularly rich. Plant products have half as much as animal products. Particularly rich plant sources are soymilk and legumes….
Q: And back to slide 6, how did you calculate the thimerosal risk?
A: I took the cumulative dose of thimerosal, 180 mg (rounded up to make it easy for comparison) and assume that’s actually a dose that’s given to 1 kg child (2 lb. child). This is a very conservative way to look at this, so if you calculated that and converted it to the same units, it would be 1 micromole per kg body weight. Again that’s conservative, its probably .1 micromole per kg of body weight. That cumulative dose of thimerosal is considerably less than that daily intake of sulfur amino acids would be and very much lower than the body glutathione pool. It’s 2,000-fold lower than the total body thiol.
Q: You say that one of the roles of glutathione is to react to chemicals in the body. Do foods also contain the reactive materials that our bodies would use glutathione to deactivate?
A: Yes. A few years ago we conducted a study to ask that question. This is in peer-reviewed literature. We measured glutathione reactive materials in foods. We looked at 140 common American foods. I list 2 to give good reference comparisons. If we look at amounts of reactive materials in an 8 oz. glass of 2% cows’ milk, that number is 21,700 nanomoles. If you convert that to micromoles and assume that child would consume 4 oz. of milk in a serving that would be equivalent to 10 micromoles. So 10 times the amount of thimerosal. Bottled unsweetened apple juice, the value of that is 6,600 nanomoles of reactive chemicals. So a 4 oz. serving is 4 micromoles. This would be substantially above the cumulative dose of thimerosal.
Q: So it takes more glutathione to react to 4 oz. glass of milk than to thimerosal containing vaccines?
Q: And the TCVs are normally administered over a 6 month period?
A: That’s correct.
Q: Are there any natural variations in glutathione levels?
A: There are a number of different variations in glutathione levels. [slide of his publication showing concentrations of glutathione in plasma, showing a variation of 30% on a daily basis in 60 people]
Q: We’ve heard about GSH/GSSG ratios, glutathione to oxidized glutathioine ratios, is that correct?
A: Yes, these are just the cysteine and glutathione concentrations themselves, but at the same time, we measured disulphide forms....It’s the reduced form that’s functional as an antioxidant. The ratio is the measurement of how good a reactant it is. These variations that you see in the reduced form really reflect the change in the redox state. So not only do you have a change in absolute concentration of the thiol, but also corresponding variation in the concentration of reduction state…. [slide 8] What those numbers show is that turnover of the system is approximately 1 micromole per kg of body weight per minute. So that upper limit of thimerosal dosing is equivalent to the amount of turnover on a per minute basis. So in one minute, there’s more thiol being turned over than the total load of thimerosal in that dose.
Q: If you were to receive a TCV, would that change that natural variation chart?
A: I wouldn’t expect it to have any effect at all. The amount wouldn’t be delivered in one minute. The total turnover is so fast, it wouldn’t have any detectable effect.
Q: So what you’re saying is that it would take less than a minute for the body to replace the glutathione that is used to bind and deactivate a thimerosal containing vaccine, even if the entire 6 month load were administered at one time?
Q: Dr. Deth relies on several in vitro studies to support his hypothesis. Do you consider in vitro studies to be reliable way to determine in vivo toxicity?
A: No, I do not.
Q: At page 4 of Dr. Deth’s report, it states that the results of in vitro studies in his lab have not been published. But the paper states, “threshold effect for thimerosal reduction of GSH is approximately .1 nanomolar indicating a remarkably potent influence on cellular redox status in human neuronal cells.” Have you reviewed the published in vitro data which measured concentration dependence of thimerosal depletion in glutathione?
A: Yes, I did. When I read Dr. Deth’s report, this particularly caught my attention because .1 nanomolar glutathione is such a remarkably low level that there’s no analytical technique I know of that would be sensitive enough to pick up that small of an effect on the glutathione system. That prompted me to go back and review the literature on dose dependence of toxicity of thimerosal in in vitro studies for comparison with the peer-reviewed literature.
Q: Special Master: Are you saying there was no way to detect that?
A: …I know of no method that could detect that low a level.
Q: And you’ve listed that literature on p. 11 and 12 of your report?
A: Yes, I was neither exhaustive nor selective. The studies seem to be relevant to me and seemed to be extremely consistent in what they showed. [brief review of these studies]
Special Master Hastings: [question about Dr. Jones’ report about inconsistency of Dr. Deth’s nanomolar results vs. literature’s micromolar results; Dr. Jones states that he is not an expert in this area]
Q: How can culture conditions determine what concentrations you need to have toxicity? A: [slide 9] We have done a lot of in vitro toxicity research and know the problems with extrapolating from in vitro to in vivo. In tissue culture, you have cells growing in a single monolayer. Above that you have culture medium, which contains nutrients for cells to grow. Different investigators will use different sizes of dish, numbers of cells and volumes of medium. Sometimes, differences are relevant, sometimes not. If you have a chemical that accumulates in the cells, then the volume of medium relative to the volume of cells becomes highly relevant….[discussion of culture volume/cells ratio]
Simply by changing the volume that you’re putting above the cells, you can change the amount of toxicant that you’re delivering to the cells. This is really a problem for comparing literature because it’s not stated how many microliters of volume were added.
Slide 10: You can see what the importance of this comparison is in terms of thiol content. If in 1 micro liter of cell content, you measured how much the thiol content would be, commonly it’s in the range of 1,000 to 10,000 micromolar that’s in those cells. That means that if indeed you put in 1 micromolar of a chemical, you’re going to be consuming a large fraction of the total thiol in the cells. And what you see in the literature, different in vitro toxicity studies is a very large number of chemicals that react with thiols causing toxicity in the micromolar range. And in a way that’s a non-specific effect because you’re putting in so much reactive material, that you’re overwhelming the thiol system. So it’s grossly out of line with what you’d see in vivo. In vivo, you have less extracellular volume than you have cellular volume.
Q: So the fewer the number of cells of the medium, the lower the toxicity threshold will be because more of the administered volume goes to each cell, is that correct?
A: Yes. The more typical dose-response curve that you would see in these experiments is at lower concentrations. You see no toxicity, but once you reach a particular range, the cells die pretty much at the same time. All the cells have the same mechanisms. They’re all responding similarly. If you tried to select conditions, instead of using 1 million cells in your dish, you could use 100,000 cells. Then you only need 1/10th as much of your toxic substance. The curve would be shifted to the left by an order of magnitude. So I cannot pass judgment on these papers in terms of this issue because none give explicit information about the volume, numbers of cells, in order to allow comparisons. Without that information, though, it’s impossible to know whether conditions were selected to affect toxicity….[clarification of prior discussion; example of things that would affect the cell culture but that are not noted in the Deth data]
When you have someone publish toxicity that is 1,000 times different from those of a dozen different papers in peer reviewed literature, that would raise questions about the differences in methodology….As a scientist, you learn to trust the scientific literature; you have to trust the published literature. When the same observations are obtained in several laboratories, you tend to give that more credibility than an unpublished report where you don’t have an understanding of why the systems are different and why the others were wrong….
Q: Based on the stark differences between published peer reviewed literature and Dr. Deth’s data, can you draw conclusions on the reliability of Dr. Deth’s data?
A: Without actually knowing the conditions in which he did it and without understanding those details, I would not put credibility in Dr. Deth’s statement at face value of sensitivity at the nanomolar concentration. It is inconsistent with the bulk of the data that I’ve looked at.
Q: One of the studies relied on by Dr. Deth is a 2005 Jill James study….Did you reach any conclusion about its relevance to what occurs in vivo with thimerosal containing vaccines?
A: I concluded that what occurs in vitro is irrelevant to in vivo toxicity because concentrations were really out of line. I went through the calculations [goes over calculations of concentration in James’ study AND the dose that child gets over the course of vaccination on slide 12]. You would have a 10,000-fold higher concentration of thimerosal in these in vitro experiments than would be relevant to the actual dosing in vivo.
Q: And this slide illustrates problems of extrapolating in vitro studies to in vivo conditions?
A: Yes. And that’s the difficulty; the total thimerosal load is simply so low as compared to what the cells see in in vitro conditions.
Q: We heard Dr. Aposhian testify that the liver contains 10 millimolars of glutathione and that is where glutathione is most concentrated. What does this graph represent?
A: The concentration here that’s given is 750 nanomoles per mg of protein. [Dr. Jones reviews calculations relevant to graph] There’s something wrong with the methodology here, and I don’t know what methodology was used, but it really puts into question all these data. The data are just inconsistent; there’s no tissue in the body that would have 20 millimolar glutathione. The glutathione concentrations here are far in excess of what tissues would have.
Q: Is it your opinion that thimerosal-containing vaccines play no significant role in sulfur metabolism?
A: Yes, I think at the doses that are included in TCVs, my understanding of what that dose is, my understanding is that it’s too low to have any significant effect on glutathione system.
Q: At doses where effects on sulfur metabolism occur, are these adverse effects?
A: [discussion of “nerf” system] The main point is that we are constantly exposed to reactive elements, so even if one sees changes, we can’t just assume they are bad changes….
Q: It’s a compensatory response?
A: Yes, it’s a protective response.
Q: And what sorts of things induce this compensatory response?
A: There are many that are common in the diet. This is the reason that your mother told you to eat your broccoli because chemicals that do this are found in these vegetables; common in other food; garlic, onion and in green tea. There are widespread chemicals in the diet that will activate this. You can activate it with a number of other mechanisms.
Q: So if I never ate my vegetables as a child, would I not have these compensatory responses?
A: No, you still have the compensatory responses, how much they are activated will vary on an individual basis. The most important thing is that simply by measuring by level and change of glutathione, you can’t assume it’s a good or a bad thing to see a change. An initial small decline can lead to an incredible increase later on. There’s a chemical in apples called dimethylfumarate that will activate this system. You’ll see a small decline in glutathione, and then glutathione will rise….
Q: So in summary, what does this data show you? At doses where you do see effects on sulfur metabolism, can you assume those are adverse effects?
A: I think the normal diurnal variation seen is greater than that which would occur due to the thimerosal dosing. One wouldn’t say that the variation is an adverse effect; it’s a normal physiologic effect. And there’s a broad range of agents that initiate the nerf 2 system so one cannot conclude that modification of glutathione implicates toxicity. That is an incorrect conclusion to make. Modification of glutathione per se cannot be taken as evidence of adverse effects.
Q: You’ve demonstrated that thimerosal containing vaccines do not significantly affect sulfur metabolism, and even if they had effects, you couldn’t assume that the effects were adverse. If there were adverse effects on sulfur metabolism, could you assume that they would be a cause of autism?
A: I went through Dr. Deth’s presentation of his hypothesis. I would like to go through briefly what I consider to be critical points, and what I felt were questions.
Q: You found critical flaws with Dr. Deth’s mechanisms, correct?
A: Yes. Here’s Dr. Deth’s slide 18 where he’s outlined his mechanism [discussion of delivery of cysteine to the neurons; Dr. Jill James’ work] What he is arguing is that you have a decline in glutathione, relevant to downstream effects, ignoring the effect that the critical, obligatory intermediate (cystanele glycene), and that number doesn’t change. That is completely inconsistent with that hypothesis. Even if you believe that plasma values are relevant, the data in the James article show that it can’t be correct according to Dr. Deth’s model.
Q: So Dr. Deth’s reliance on the data refutes his hypothesis?
A: Yes, this part of his hypothesis really cannot be. The second point on this slide: Dr. Deth relies upon an inhibition of cysteine transport as particularly critical in the mechanism.
It’s important to introduce the discussion of the way the body maintains amino acid supply. Slide 22 illustrates this basic cell physiology. The way that cells make proteins, proteins require 20 amino acids. So the question is, how does the cell avoid only having 19 of those? How do all of the cells assure that all have all 20 amino acids? They do this by maintaining multiple amino acid transporters. [discussion of amino acid transport] What these data show is that over a broad range of thimerosal concentrations, there’s essentially no effect on cysteine uptake. [discussion of cysteine uptake] So this hypothesis carries very little weight. This component of the hypothesis is also incorrect. His data doesn’t show that you would have a sufficient inhibition of the cysteine uptake to have any effect on the downstream pathways, as he has argued.
There’s a real question with regard to Dr. Deth’s argument as far as what determines if homocysteine goes through the degraded pathway or recycling pathway of methylation. (explains chart on slide, hard to understand without graph). So the question is what happens to the homocysteine?....[importance of methionine and S-adenosyl methionine (SAM)]
Q: And you found one more critical flaw in Dr. Deth’s hypothesis, is that correct?
A: [slide 25] Yes, the critical flaw concerns step 4, closing the cycle in terms of signaling involves methionine synthase reductase….If we go back to Dr. Deth’s mechanism, it’s completely inconsistent, the data show that it’s protective. At four critical sites in this scheme, the pieces just don’t fit together for changes related to sulfur amino acids as a cause of autism.
Q: So the data Dr. Deth presented in formulating his hypothesis do not support the hypothesis?
A: Yes; these are the data he presented. What the data show to me is that there really is not appropriate evidence, there is just not the data that the dose of thimerosal can alter sulfur metabolism. This is not established by the data…. There really is no plausibility to this hypothesis at all. There’s a “scaffold without a building;” there are a lot of components to it, but it doesn’t have the strength or solidity to be reasonable or plausible science.
In terms of summary comments, (1) the cumulative dose of thimerosal is too low to cause the cumulative effects on glutathione metabolism. That’s not plausible. (2) The natural variations in glutathione are greater than what one would expect from thimerosal. (3) If you did have an effect, you would have to conclude that the low dose would be protective. (4) And in vitro studies show that thimerosal disruption is occurring under nonspecific conditions. Cellular conditions are set up in a way to disrupt lots of things because they are in irrelevant concentrations. I have to conclude that the data do not support this hypothesis.
Cross Examination of Dr. Jones
Q: [Mr. Williams] I want to ask you about a paper you published about mercury toxicity. This wasn’t mentioned in your direct testimony, but it’s on your CV and you’re probably familiar with this paper?
Q: Let’s show the title: Differential Oxidation of Thioredoxin 1, 2 and Glutathione by
Metal Ions. You published quite a few papers on thioredoxin?
Q: What is it, and what is its significance in the brain?
A: I don’t concentrate on that. We study mostly cellular mechanisms. These are studies to understand the thioredoxin system in cells….
Q: How do thioredoxins in the cell or mitochondria relate to glutathione?
A: They are complementary systems; glutathione is not alone. There are many antioxidant systems.
Q: This is one of your original research papers?
Q: And you looked at quite a few metals but one of them was mercuric mercury?
Q: Why did you look at mercury and thioredoxins?
A: We were looking at different metals to see if some affected thioredoxin differently than they affected glutathione; it was a comparative study.
Q: How was it funded?
A: This was funded by my NIH grant; it’s an oversight if it wasn’t cited.
Q: If mercury were able to significantly inhibit these two enzymes (thioredoxins), would that be a sign that it’d be toxic to cells?
A: Certainly, we used very high concentration, but what the data showed is that you can in fact change the function of thioredoxin by putting in mercury.
Q: This is both in cells and mitochondria. If you had enough mercury there to suppress or inhibit this enzyme, would this happen in both mitochondria and cells themselves?
A: Yes, it would happen where mercury accumulated. When we use these cell culture studies, we were not addressing in vivo relevance of specific dosing. We were addressing scientifically whether the thioredoxin system responds in the same way as the glutathione system, and the answer is no – they respond differently.
Q: If we turn to the second page of this paper --you’re talking about discrimination between oxidized and reduced glutathione and the thioredoxin system. These would be in brain cells, too, right?
Q: This discrimination between the glutathione and thioredoxin systems could be very informative of mechanisms of toxicity. Right?
A: We have multiple protective systems. We have to think in terms of global imbalance of pro-oxidant and anti-oxidant systems….
Q: If a toxin could inhibit the thioredoxin system but didn’t affect glutathione, would this mean that it was harmless?
A: The systems are complementary. If you affect one of the systems, the other system takes over. They often have a difference in sensitivity, which is what we were investigating here.
Q: If you had a significant inhibition of thioredoxin, would that be harmless because of compensatory mechanisms?
A: I can’t really answer that question directly. There are drugs that have been developed targeting thioredoxin to use in anti-cancer therapy to try to target and kill the cancer by specifically inactivating the thioredoxin system. The concept is out there in the scientific literature.
Q: The concept is that you could have a toxic effect just by affecting the thioredoxin system?
A: Yes, if you could find something that would do that.
Q: Would you explain what nrf-2 is?
A: If you have something that gives you a low level of oxidation or challenge to the cell, that would potentially perturb glutathione system, that turns on the nrf-2 system, by turning it on, it causes an increase in the thioredoxin system. Both of these systems are responding together.
Q: [Mr. Williams quotes again from a paper.] What is ask-1?
A: It’s an enzyme that can activate a cell death program.
Q: And that’s the concept behind these cancer therapies?
Q: You found in this study that mercury has different effects than some of the other metals?
Q: What did mercury affect in your study?
A: At 10 micromolar and 100 micromolar mercury, it caused a change in the thioredoxin and activation of the ask-1 mechanism. [quotation from paper] In contrast to other metals, mercury was oxidized by thioredoxin enzymes but had little effect on the glutathione cycle?
Q: And it did induce cell toxicity after 24 hours. You talked about how you have to give it enough it. You think this was long enough?
A: Yes, with a high enough dose. In general, it really depends on cell type. When you take a cell line and put them in culture. We had a system where we knew we could kill the cells under these conditions, and we wanted to see what the effect was on the thioredoxin and glutathione systems.
Q: You only used 10 micromolar?
A: Yes, we only used 2 conditions.
Q: Do you know what would happen if you used 1 micromolar of mercuric mercury?
A: The study was really designed by Jason Hansen. He took this from the literature and in his analysis, we needed 10 micromolar in order to kill these cells.
Q: Before the cell dies, wouldn’t it become dysfunctional, for a few of those hours?
A: We didn’t really study anything else.
Q: So you were looking for cell death as an end point of the study, not dysfunction?
Q: Isn’t it reasonable to think that a lower dose would cause dysfunction before it caused death?
A: I can’t speculate on that outcome; we tested a hypothesis about cell death.
Q: You discovered a difference between effects on mitochondria and the cell itself, didn’t you? [quoting from Dr. Jones’ paper] “…mercury has greater oxidative effects on the mitochondrial compartment than the cytoplasmic compartment.”
A: Yes, that’s correct….
Q: In this instance, oxidation is bad for the cell?
A: That is a question that we really don’t know the answer to. In human bodies, even under normal conditions, cells are more oxidized than under these laboratory conditions with toxic metals.
Q: Was the effect of mercuric mercury enough to kill the cells?
A: In this model, we don’t even know that. What we know is that these cells under these conditions with a high amount of mercury died. We can’t say that this is the cause of death in these cells. I can see why you would want to extend that into a logical argument, but it’s an association at this point, not a causal relationship. You can say mercury caused cells to die, but not because of change in thioredoxin.
Q: Why does NIH spend money on this?
A: Because we are the pioneers in understanding the different cellular compartments. We’re learning how to regulate different compartments in the cells.
Q: This is where you show relative effect of different metal ions, figure 4. There are 3 of them that are particular high, and one of them is mercuric mercury?
Q: What are you measuring here, the relative effects of the metal ions?
A: Yes, but remember that these are at 100 micromolar mercury: this is a horrendously high concentration. These are an indirect assay of the ask-1 activity, one of the multiple cell death activating pathways; these three metals give more activation of that enzyme.
Q: You’ve made a point all morning that even though we see these toxic effects of mercury on these enzymes, the doses are so high that they’re not relevant. In the monkey studies, the levels were at 30 nanomolars.
A: I’d have to go back and check, but yes, it was a very low amount.
Q: In your report you make a comment about the Burbacher paper; you said in your report that this primate model (the infant monkey study), you say it showed clear evidence for delivery of mercury to the brain?
A: I don’t see any way you could avoid this conclusion; that’s what the data show.
Q: But it was delivered in a dose that was much lower than the dose you measured?
A: There is a difference between being able to detect something and having a concentration that’s relevant to the in vitro study. The ability to detect has to do with sensitivity. Ability to detect something at very low levels is not really the relevant question. The relevant question is how much.
Q: Let’s talk about your background. You spent quite a bit of time at Karolinska Institute. How many times have you been there?
A: In 77 and 78, I was back in 80 or 81, 82 or 83, 87, 94 and 97 for varying periods of time.
Q: Do you know who Marie Vadder is? She’s the author on the adult monkey study from Karolinska Institut.
A: No, I don’t know her.
Q: But you do know Arne Holmgren?
Q: How long have you known him?
A: Perhaps 30 years.
Q: You published in his books?
Q: You respect his work?
Q: Do you have a way of tracking when your papers get cited?
A: I really don’t spend my time on that; I spend my time on original research.
Q: We found a paper by Holmgren that discusses your paper. And this is the same thioredoxin system that you studied?
Q: This is a study of mercurial mercury on thioredoxin. You haven’t seen this paper before?
A: No, I haven’t, so I would need quite some time to study it before I can comment.
Q: If I ask you a question you can’t answer, that’s OK but let’s do the best we can. It does cite two of your papers, including the one we talked about. [quotation from Holmgren paper; discussion of paper specifics] ….Well, let’s see what Dr. Holmgren says: “Overall, mercury inhibition was selective toward the thioredoxin system, in particular, the remarkable potency of the mercury compounds to bind to the …thiol on the active side of thioredoxin reductase should be a major molecular mechanism of mercury toxicity.” You don’t agree that this is a way to interpret this paper?
A: Dr. Holmgren is a biochemist who works with purified proteins, and as far as pure chemistry, he does a first rate job of that. He does some cell culture work, but he has not done in vivo work. These are studies aimed at understanding molecular mechanisms. My feeling is that you’re really going after something that’s irrelevant to human exposure when you start trying to extrapolate from something that’s even worse than a cell study; it’s a purified protein study….
Q: [Questions about a graph in a paper Dr. Jones hasn’t read] Is the graph measuring nanomolar concentrations?
OBJECTION: The witness hasn’t had time to read the article; he needs time to review it. We should have had notice of this article.
Special Master: We can proceed to ask questions about it, but if you continue getting answers that “I don’t know”, it’s not much use to us up here, and again, that’s your audience.
Q: I’ll make this last point, the last page says “we have demonstrated that in a cellular system, the inorganic form of mercury preferentially targets the thioredoxin system, of which thioredoxin reductase was inhibited to a greater extent than thioredoxin itself by more potent mercury concentrations.” So they weren’t studying just isolated proteins?
A: I have not looked at the data, I have not read this paper.
Q: What function do mitochondria play in a cell?
A: Mitochondria are the energy source, the major supplier of ATP, major energy currency of cells.
Q: You have published a lot on mitochondria? You’re one of the leading experts?
A: I have expertise, yes.
Q: I want to show you a case report on a child with mitochondrial dysfunction and autism.
OBJECTION: I’m objecting because we’re going to get to theory 2(c). There was nothing in Dr. Deth’s report on effect of thimerosal on mitochondrial function.
Special Master: Let’s see where we’re going to go with it.
Q: Did the DOJ lawyers show you this paper?
A: No, my expertise is sulfur metabolism and oxidative stress; I haven’t seen this paper.
Special Master: You included on your slide 20, the table of results from the James study involving various transmethylation and transsulfuration metabolites. You highlighted cystenele glycine. You did not comment on those levels that Dr. James found that were far lower in autistic than control samples. Do those have any bearing on anything you said?
A: The most difficult aspect of this type of comparison is that in almost every disease population, the glutathione levels are different than controls.
Q: Any disease?
A: Pretty much. Cardiovascular, diabetes, liver, lung disease. There are general differences, and most commonly, reduction in glutathione.
Q: So the reduced level is a response to a disease process?
A: Right, it seems to be more a response rather than a cause. It’s a common consequence of different diseases that we have studied.
Direct Examination of Dr. Kemper
Dr. Kemper is a neurologist and professor at Boston University School of Medicine and now works primarily in brain research. He has published about 170 articles, including 30 on autism. He serves on several editorial boards of journals and has reviewed several hundred articles.
Q: What is neuropathology?
A: It’s the study of diseased brains, nerves and muscle.
Q: Where do neuropathologists obtain samples to study?
A: For routine pathological studies, autopsies and surgical specimens. For my own work, we use brain banks.
Q: What is a brain bank?
A: They are established by the federal government; they receive brains, process them in a uniform matter and make them available to researchers.
Q: Do brains come from organ donors?
A: A lot of people just want to have their brains examined, especially in autistic people. They just donate them.
Q: What is the goal of neuropathology?
A: For routine work, it’s diagnosis which dictates treatment. For my work in anatomy of various diseases, I try to find out what is the organization of disease, the nature of disease. It creates a gold standard for other investigators, so whatever they find has to explain what is obvious from morphology.
Q: So would it be fair to say that neuropathology is important to both diagnosis and its cause?
Q: When did you begin researching brains of autistic individuals?
A: First brain we got in 1980. It was an extremely well documented brain; the patient had been seen by everyone in Boston, so there was no question about diagnosis. We had the technique to do whole brain serial sections, so we could compare every part of the brain to controls. We published the data in 1985.
Q: You said we?
A: Dr. Margaret Bowman and I; we work together.
Q: Would it be fair to say you and Dr. Bowman are pioneers in researching brains of autistic individuals?
A: Before we did this anatomy, it was believed that autism was caused by parental behavior. Guilt trips were laid on these people. It was comforting to me and to Dr. Bowman to find a structure.
Q: When did you publish your first article on this topic?
Q: What was the response to that?
A: It was amazing. The interest was strong, and it was on the front page of Boston Sunday Globe.
Q: Would it be fair to say that you dedicated a large portion of your life to investigating neuropathogenesis of autism?
Q: Can you comment on the number of brains that have been studied in this area?
A: Actually, relatively few. There have been problems with availability and people with interest to study them.
Q: Despite the relatively small number, has there been a consistency among the findings?
A: It’s rather amazing; even reports from different labs are showing consistent findings.
Q: What is the age of the brains that have been studied?
A: Autism isn’t diagnosable until age 2. The earliest ones available are 3 years of age. In our own studies, we start at age 4 and go up to age 50.
Q: And that’s because people don’t typically die of autism?
A: Not from, but because of autism. They have seizures; there are drownings; appendicitis when the individuals couldn’t tell their parents about the pain.
Q: You mention that there have been consistent findings. Are the findings consistent across age ranges?
Q: Are you or other researchers able to determine if brains came from regressive or classic autistic individuals?
A: In our field, that information is only available on one study and that’s the Vargas study, finding no difference in morphology between regressive and nonregressive autism.
Q: Have you seen anything that leads you to believe that the brains are different?
Q: Based on your own research and review of the literature, are there structural changes in the brains of individuals with autism that suggest problems in brain development?
A: Yes; there are plenty of them.
Q: Are they in specific areas of the brain associated with autism?
A: There are changes in the brain stem, medulla, cerebellum, in the cerebral cortex and in the limbic system.
Q: Are the findings in the literature found universally in every brain?
Q: Can you explain what the findings have been and relevance in terms of consistency of the findings?
A: There’s a lot of pathology in almost all of the brains, and consistent areas that seem to be involved. There are eight brains presented here from age 10 to 45. Only the cerebral cortex was studied here in three brain areas. Bailey shows similar pathological changes and additionally studied brain stem and cerebellum. He lists brain abnormalities in neural migration in 4 of the 6 and Purkinje cell abnormality, which is the most common difference found in autism.
Q: What is the significance of the various findings that are noted in this slide in terms of brain development?
A: This slide is a timeline of developmental events in the human brain. [outlines what other experts will testify on in brain development and gives an overview of development of brain]
Q: Now that we have an overview, let’s talk about some of the specific findings on these slides. Can you tell us what part of the brain this is?
A: This is the medulla, the bottom part of brain stem in an embryo; this is the germinal zone.
Q: And what is this?
A: The germinal zone makes neurons. They are born in one spot and migrate to another and settle in. If this process is arrested at any point, you have a good idea of timing of the malformation. [discusses specific migration patterns]
Q: So this slide shows normal migration?
A: Yes. Now slide 6 shows one of our own personal cases here. Arrows point at neurons arrested in the rhombic lip germinal zone.
Q: How can we tell that there’s been an arrest in migration?
A: You can’t, but I can under the microscope by seeing that these are neurons.
Q: This is a brain that you personally studied?
A: Yes. And the Bailey paper had neurons arrested along here, but there’s no good illustrations of that.
Q: Please comment on slide 7.
A: This explains what’s in the slide, picture of neurons in an autistic brain.
Q: It appears that the normal brain and autistic brain are different shapes?
A: The angle of cut is slightly different on those two brains.
Q: So we’re still looking at the same area?
Q: Slide 8.
A: Here’s the inferior olive, loaded with neurons. You can see on the slide that the neurons are lined up in a row. That’s an abnormality. Contrast it with the picture of normal appearance of that region. This is a fairly consistent abnormality in autism. It indicates abnormal settling in of neurons that have migrated from the rhombic lip.
Q: Slides 6-8 that deal with migration from the rhombic lip to the inferior olive, can you correlate those findings to a specific period of brain development?
A: Yes, these are probably from about 14-16 weeks of gestation.
Q: Slide 9, what area of the brain is this?
A: The cerebellum. This illustration is from our original case, what we’re showing here is a loss of Purkinje cells, and a loss of granule cells where loss of Purkinje cells is particularly profound.
Q: And what is the Purkinje cell?
A: It’s the boss cell of cerebral cells.
Q: Quotes from p. 18 of Dr. Kinsbourne’s report. Does this statement indicate that Dr. Kinsbourne is comparing parametal cells to Purkinje cells?
A: It’s very difficult to compare them. They are very different. [looks at slide; points out loss of Purkinje cells in autism] This shows mild loss with relative preservation of granule cells. Here, in a different picture, you see profound loss of Purkinje cells with attendant loss of granule cells.
Q: How can you tell?
A: By comparing densities. In the autistic brain, this pathology is in the more lateral part of the cerebellum. In the vermis, it is spared in terms of cell loss. This will be important later on.
Q: Findings in terms of Purkinje cells are the most consistent in autistic brains?
Q: And you can date the loss of Purkinje cells to a specific period of brain development?
A: We think we can. Slide 10 illustrates that the inferior olive usually projects… Purkinje cells. These climbing fibers from the inferior olive reach the Purkinje cells at about 30 weeks of gestation. Prior to that time, it’s not involved with the Purkinje cells. Since we know that loss of Purkinje cells is from birth out, so the normal connectivity between these climbing fibers and the Purkinje cells was not in place at that time.
Q: The climbing fibers reach the Purkinje cells around 30 weeks of gestation?
Q: In normal brains, if Purkinje cells are lost after they were formed, the inferior olivary neuron goes away as well. And in autistic brains, the Purkinje cells are not there, but the inferior olivary neurons are still there. So you conclude that Purkinje cells were gone before the relationship between the inferior olivary neurons and Purkinje cells was formed?
A: That’s our best guess. [discussion of brain pictures on slide]
Q: Slide 11. What does this one show?
A: This is an illustration from Pashko Rakesh in 1982. This shows what the cerebral cortex looks like at 9 weeks of gestation. The next images are from 13, 16, 21, 25, 30 weeks of gestation. You see the climbing fibers there at 25 weeks of gestation. In the next five weeks, they envelop the Purkinje cells. By birth, they envelop them as densely as we saw in the prior slide.
Q: So at 30 weeks of gestation, the climbing fibers reached the Purkinje cells?
A: Yes, and that’s shown in other studies as well.
Q: Please talk about slide 12.
A: I want to talk about large neurons. You can see how large those neurons are in the autistic brain. They are unusually large. In the older brains, over 17 years of age, you can see that the cells are now small.
Q: Slides 12 and 13; what do those changes signify?
A: It’s difficult to interpret. Enlargement of neurons is very unusual; there’s very little literature on it. The majority of studies has to do with the inferior olive. Lesions lead to abnormally large neurons; so it’s a disturbance in connectivity. This means that the circuitry in the cerebellum is not normal.
Q: So these slides show disturbances in brain development?
A: Yes, in connectivity.
Q: At what point in development do the changes that we just saw relate to?
A: Childhood up to adulthood.
Q: Let’s look at slide 14.
A: This is development of the cerebral cortex. In 1970, a whole new concept had appeared in the development of cerebral cortex. [Discusses various panels on slide] [Describes the sudden split into two zones of the cerebellum.] The bottom zone is subplate. Top zone is layer 1 of cerebral cortex. Between are definitive neurons of definitive cerebral cortex. These neurons are born between these two zones. The reason that’s important is that the zone persists up until just after birth and then disappears. Experimental studies show that the subplate is important for the establishment of circuitry in the cerebral cortex.
Q: Slide 15?
A: This is to show there are no good examples in the neuropathology literature of neurons arrested at the germinal zone for the cerebral cortex or arrested in the migration. Examples that are available in the autistic brain are examples of abnormal settling in in the cerebral cortex. This is an abnormally folded cortex. 3-4 times as many folds as one would expect. Too many small gyri. Pathogenesis has not been well worked out. Some kind of destructive process in the cerebral cortex around the end of migration of neurons into this area. Bailey’s illustration shows disoriented pyramids in the cerebral cortex. There is marked abnormality in the orientation of these neurons.
Q: How are they normally oriented?
A: Straight up. Something happened in the settling in.
Q: Where are the neurons supposed to be?
A: Towards the top of the picture.
Q: Does this indicate that the neurons stopped during their migration?
A: That would be the interpretation.
Q: Slide 16?
A: This is a fairly striking malformation in another autistic brain. This is the anterior singular gyrus.
Q: You’re point to the blue swirl?
A: Yes. [discussion of brain pictures on slide]
Q: Slide 17?
A: In all autopsy series, mainly in Hutzler and Bailey, they reported increased numbers of neurons in the white matter in the so-called subplate zone.
Q: And at the top there are dark blue dots and at the bottom lighter blue?
A: These neurons are not very large; in fact, there are neurons in the white matter.
Compared to the control brain, you see only occasional neurons in that region. Another thing is demarcation between the cortex and the white matter is not as striking as it is normally; this is another change that has been reported in the autistic brain. These subplate neurons have been noted in numerous brains, including in the schizophrenic brain as well, and that zone is a gigantic synaptic zone in development, and these cells should disappear shortly after birth.
Q: Slide 18?
A: This is from Hutzler. This is the surface of the brain. This is a cell layer of cortex and layer 1. There are large numbers of neurons here; multiple times what there should be.
Q: So slides 15-18 are malformations that have been reported during development?
Special Master: Are you saying that there’s an increased number of neurons in the brain?
A: Only in the white matter, just below the cerebral cortex.
Q: Are they in the wrong place or too many?
A: Too many.
Q: What period in brain development are these changes in the cortex associated with?
A: During the settling in of neurons into the cortical plate, somewhere between 16-20 weeks of gestation.
Q: Slide 19. What area of the brain is this?
A: This is the hippocampus. This is the memory circuit. The cells are classically divided into fields. The memory circuits are sequential from one cell group to another; it’s called the trisynaptic circuit. At the top is the autistic brain, at the bottom is the control brain. [contrasts autistic and normal brains] Cells appear more tightly packed together in the autistic brain….The next slide is a more highly powered view of these two fields.
Q: And can you relate these changes to a specific period of brain development?
A: No; these are the ones we can’t tie to a time. Timing depends on migrations and synaptic hook ups.
Q: Slide 21?
A: [explains differences in limbic system]
Q: Before we move on, we’ve gone through some examples of changes in the autistic brain. Would it be fair to say that the structural changes we’ve reviewed most likely occur prenatally?
A: The majority of them, yes.
Q: And one of your slides indicated a change in subplate neurons that occurs just after birth?
Q: Is Dr. Kinsbourne proposing a mechanism by which thimerosal from vaccines interferes with neuronal development?
Q: Is Dr. Kinsbourne proposing a mechanism that involves any of the developmental processes you discussed here today?
Q: Another of Petitioners’ experts, Dr. Aposhian, listed six pillars that he believes support conclusions that thimerosal vaccines contribute to autism. The sixth pillar was based on the Courchesne article. Dr. Aposhian says it provides evidence of postnatal loss of brain cells in the cerebellum in autistic individuals. Have you reviewed that article?
Q: Is that how you interpret it?
A: No. If our interpretation is correct, then they were lost early, and there is no reason for them to form again.
Q: Are they talking about the decrease in number of Purkinje cells?
A: That’s right.
Q: Is your opinion that there’s not a loss of Purkinje cells based on the notion that Purkinje cells just were not there to begin with?
A: It’s hard to know whether they were there to begin with or not.
Q: But in any event they were lost before they established relationship with the climbing?
Q: You also addressed in slide 22 the issue of abnormal postnatal brain growth.
A: I selected a few articles to show; here’s one from Dr. Courchesne. [Discusses illustration on slides] Right after birth, there is a gigantic burst of brain growth. According to Dr. Dawson’s paper, that spurt of growth may be confined to the first year of life. And then brain growth slows, and almost comes to a standstill, whereas the normal brain continues to grow. At adolescence, the brain size is the same between autistics and controls. So there’s a peculiar feature of the autistic brain. [discussion of various studies of brain growth]
Q: So, in short, the studies are showing there is rapid and significant brain growth in the first few months of life in these autistic brains?
A: Yes. [discussion of head size study among autistic individuals showing uniform fast head growth in the early months]
Q: Has anyone come up with a good theory that explains abnormal postnatal head growth?
A: There are a lot of ideas, but I don’t see any of them that really work. One of the more popular ones is Dr. Casanova’s, which you’ll hear in a couple days, about an increased number of brain columns. I went through the entire literature and one thing that showed up is in the dyslexic brain, where there are also malformations in the cortical plate, they are called brain warts. It was found that there is increased local connectivity but decreased colossal connectivity. The timing of synaptic elimination doesn’t fit. It’s really a mystery.
Q: In your opinion, is abnormal postnatal brain growth consistent with exposure to ethylmercury from TCVs?
A: No, that generally makes brains smaller.
Q: Have you reviewed the literature that reviews the neuropathology of mercury toxicity?
A: Yes. [slide 23]
Q: What did you find?
A: These are summary slides. Pathology is very consistent from report to report. [examples from Minimata mercury toxicity studies] There are a lot of issues in the visual cortex; it involves the peripheral part. A common complaint of people with mercury toxicity is tunnel vision, and eventually blindness. The motor cortex and sensory cortex are included. The auditory cortex is also implicated; deafness has been reported. Also the cerebellum has been almost universally involved in mercury toxicity. So, there are destructive legions, neurons are killed by mercury toxicity in the motor-sensory cortex, the visual cortex, the auditory cortex and the cerebellum. It’s a very destructive process.
Q: Let’s go to slide 24.
A: All the literature points out this anomaly in mercury toxicity; the pathology involves the deeper parts of cerebellum. The other point is that this is the vermis, the midline of the cerebellum; this is the area of predilection for mercury toxicity. In the autistic brain, it’s different; it’s in the lateral lobes. The other feature which has been brought up in all the papers are the Purkinje cells; there are almost no granular cells. There’s a striking predilection for granule cells. The autistic brain is just the reverse of this.
Q: Let’s look at slide 25.
A: This is from Shirate. It uses myelin stain. This is the visual cortex; it should look like this but it doesn’t. It is severely destroyed in mercury toxicity. Myelin is gone, the cortical plate is gone….
Q: Based on your review of the literature and your knowledge of neuropathology of autism and mercury toxicity, do you have an opinion as to whether or not they’re consistent?
A: No, they are not consistent.
Q: Can you summarize the differences between the neuropathology of mercury toxicity vs. the neuropathology of autism?
A: [slide 26] The initial clinical feature of mercury toxicity is a sensory neuropathy – a numbness and tingling in their feet. This is not a clinical feature of autism. A restriction of the visual field is a classic finding of mercury toxicity; it is not found in autism. Ataxia and and ...are the descriptions for deficits in the cerebellar region; they are prominent features from mercury toxicity, and there’s no trace of it in the clinical picture of autism. In neuropathology, there’s no evidence of abnormal brain growth in mercury toxicity; in many reports there are small brains. In autism, no evidence of involvement of peripheral nerve on autopsy. Hippocamus involvement in autism, but spared in mercury toxicity. [further review of the slide, mentioning loss of Purkinje cells in autism and preservation of Purkinje cells in mercury toxicity; secondary loss of granule cells in autism; primary loss of granule cells in mercury toxicity; predilection for lateral lobes in autism vs. midline for mercury toxicity…neurons, visual cortex differences]
Q: Petitioners have presented a hypothesis that rather than a direct toxic effect, it is persistent inorganic mercury in the brain causing inflammation that leads to autism. They’ve relied primarily on work by Vargas to support this hypothesis. Have you reviewed the work of Vargas, Zimmerman and Pardo?
Q: Do you personally know Drs. Zimmerman and Pardo?
A: Yes, I do; particularly Zimmerman.
Q: Have you discussed their work with them?
A: Yes, I have.
Q: Did you inform us that you spoke with Dr. Pardo about his work?
A: Yes I did.
Q: Can you describe that conversation?
A: I met him at a meeting we had convened for autism. He gave a paper there on his work on the immune system. I was particularly interested to hear his views on what he thought the relationship was between the neuroimmune response and the pathological changes we had seen in the autistic brain.
Q: During that meeting, did you recommend that we speak to Dr. Pardo?
Q: Did you recently receive a letter from Dr. Pardo?
Q: And are the statements contained in Dr. Pardo’s letter consistent with the discussion you had with him?
Q: What is microglial activation?
A: That means that the glial cells are more prominent within the tissue in general; the nucleus may be enlarged….
Q: On page 13 of his report, Dr. Kinsbourne lists the characteristics of neuroinflammation as follows: “edema, activation of microglia, and local invasion of immune cells from the circulation.”
Q: Would you agree that this is an accurate description of neural inflammation?
A: Only the glial cell response.
Q: Did Dr. Pardo and his colleagues indicate anywhere in their publications that edema is present?
A: No, and we found no evidence of this either.
Q: Did Dr. Pardo and his colleagues indicate that local invasion of immune cells from the circulation occurs in neuroinflammation?
Q: Would that be consistent with Dr. Pardo’s findings?
A: Yes, he reported the lack of an adaptive immune response.
Q: So the local invasion of immune cells refers to an adaptive immune response?
Q: So do you believe that Dr. Kinsbourne’s description of neuroinflammation is incorrect?
Q; Is microglial activation present during normal brain development?
A: Yes it is.
Q: So microglial activation is not specific to the presence of a neurotoxin, such as mercury?
Q: Does Dr. Pardo say whether it’s possible that developmental abnormalities present since gestation could produce microglial activation?
A: Yes he did.
Q: And does Dr. Pardo state that prenatal neurodevelopmental abnormalities are consistent with neuroinflammation?
Q: Does the Vargas paper also mention the possibility that activated microglia reflect continued patterns of abnormal development that began prenatally?
Q: Can microglial activation be a beneficial process?
Q: Did the Vargas paper discuss the idea that microglial activation can act as a neuroprotectant?
A: Yes, it was a major point.
Q: So could microglial activation be a response to a disease rather than its cause?
Q: Based on your review of Dr. Pardo’s work and his letter, does he assume that microglial activation is causing autistic symptoms?
Q: Now I want to turn to astrocytes. At p. 17 of Dr. Kinsbourne’s report, he states that “activated microglia can kill astrocytes with friendly fire.” Does this statement suggest that part of Dr. Kinsbourne’s hypothesis is that astrocytes are dying?
Q: And Dr. Kinsbourne calls gliosis “the sequel of the death of astrocytes and inflammation.” Is that an accurate description of gliosis?
Q: What is gliosis?
A: Gliosis is the enlargement of glial cell bodies, including their nucleus and cytoplasm, and they are more easily stained.
Q: Do you see astrocyte death with gliosis?
Q: So when on p. 13, Dr. Kinsbourne says the presence of gliosis provides “dramatic support for his hypothesis,” do you agree with that?
Q: Based on your review of Vargas’ work, did they see increased or decreased astrocyte activation in autistic brains?
Q: Is this finding consistent with Dr. Kinsbourne’s hypothesis of astrocyte death?
A: No. There was no death reported there.
Q: Based on Vargas’ work and Dr. Pardo and Dr. Pardo’s letter, is it your opinion that astrocyte activation is consistent with Dr. Kinsbourne’s suggestion of astrocyte death?
Q: When he testified, Dr. Kinsbourne stated that astrocyte death may not be necessary to his model, but astrocyte inactivation or malfunction would be enough. Is increased astrocyte activation, as shown in the Vargas study, consistent with malfunction of astrocytes that Dr. Kinsbourne testified about?
A: Not that I know of.
Q: Would you say based on your review of the Vargas work, their findings are inconsistent with Kinsbourne’s statement that astrocytes are dead or no longer active in autistic brains?
Q: Did the Vargas study try to correlate neuroinflammation to regressive autism?
A: No. As a matter of fact, it made a point that several patients had regressive autism, and there was no difference in the immune response of regressive and nonregressive patients.
Q: So they did look for correlation?
A: They did.
Q: And they found no correlation in neuroinflammation?
A: That’s correct.
Q: Did Drs. Pardo, Zimmerman and Vargas conclude anywhere in their articles that neuroinflammation is the cause of autism?
Q: I’d now like to ask you about the Lopez Hertado article that Petitioners have submitted. Did you review it?
Q: Do you know what journal that article was published it?
A: The American Journal of Biochemistry and Biotechnology.
Q: Have you ever read an article in that journal?
Q: Did you try to look for it on your own?
Q: Where did you look?
A: The entire Harvard Medical School library system.
Q: Did you look anywhere else?
Q: Did you find it?
Q: Does the absence of the journal at the Harvard Library indicate anything to you about the significance of this journal?
A: This is second largest library in the country, and for some reason it didn’t have it.
Q: In the Lopez Hurtado article, what areas of the brain did the researchers look at?
A: They looked at speech-related areas [names and numbers of areas].
Q: Why would these areas be of interest to these researchers?
A: Because of the involvement of language dysfunction in autism.
Q: What results are reported in this study?
A: They thought there was a decreased number of neurons, increase of number of glial cells, and accelerated lipofuscan accumulation.
Q: Did you review their methodology?
Q: Were there any problems?
A: There were some problems that were not addressed.
Q: Could you give examples, please?
A: There’s a lack of definition of the brain areas. Without highly specific definitions, you can’t be sure that these measurements are all in the same place. The other problem is that the pigment varies from one area to another; there’s no assurance in the paper that they had done that.
Q: Was there a problem with cell counting methods?
A: Yes, that method would not have been accepted in any of our critically referred journals.
Q: Did this study look at microglia?
Q: How can you tell that it didn’t?
A: It didn’t mention it.
Q: What did they stain for?
A: They stained for standard stains, for which they measured neuron densities, and possibly glial cells and they stained for lipofuscan pigment.
Q: So when this paper refers to glia in the results section, is it fair to say they may be referring to astrocytes when they say glia?
A: I would presume so by the way it’s written.
Q: So this study doesn’t really tell us about the microglial activation in the Vargas study?
Q: This paper reports decreased neuron density in two brain areas. Do you believe the methods used were appropriate for assessing neuronal density?
A: No, I do not.
Q: Given the flaws in the cell counting method you discussed, how much confidence would you have in the results of this paper?
A: It’s a very interesting idea, and I would like to see it done with proper technique.
Q: This paper also reports increased density of astrocytes. Would that finding, if it’s correct, be compatible with astrocyte death?
Q: In your opinion, would this paper be accepted in a reputable journal?
Q: Does it add to the work of Vargas, Pardo and Zimmerman?
Q: Would you personally rely on this study for neuropathological findings?
A: No, I’d be very careful.
Q: I want to summarize your opinions about Dr. Kinsbourne’s neuroinflammation hypothesis. Do you believe microglia are damaging the brain in autism?
Q: What’s you main reason for that belief?
A: Its role in the neuroimmune response which, according to them, is very inconsistent with the widespread defects in brain development we had noticed.
Q: Do you believe it more likely than not that microglia are destroying astrocytes in autistic individuals?
A: No, there is no good, credible evidence for the loss.
Q: Do you believe it’s more likely than not that astrocytes are dying in autistic individuals?
A: I don’t believe that.
Q: Do you believe it’s more likely than not that astrocytes are inactive in autistic individuals?
A: The literature is quite the reverse.
Q: Do you believe that the roles that microglia and astrocytes play in autism are more likely than not explained by prenatal factors?
Q: Do you believe that Dr. Pardo’s work supports Dr. Kinsbourne’s mechanism of postnatal neuroinflammation as the cause of autism?
Q: Do you believe that neuroinflammation is a likely explanation for any of the structural changes you and others have observed in the brains of autistic individuals?
Q: Do you believe that it is more likely than not that any neuroinflammation that might be present in the brains of autistics is a response to the developmental abnormalities you and others have observed in the brains of autistics?
Q: Do you hold these opinions to a reasonable degree of scientific certainty?
A: Yes I do.
Cross Examination of Dr. Kemper
Q: [Mr. Powers] You testified that neuropathology is important both to understanding the etiology of autism and to diagnosing autism. Can you explain what neuropathological findings are currently used to diagnose ASD?
A: Do you mean techniques?
Respondent’s Objection: That’s not what his statement was.
Special Master: Go ahead and ask the question and see if you can get an answer.
Q: You were asked a question if neuropathology was helpful in both “determining the etiology and diagnosis of autism.” So my question to you is, what neuropathological findings are used to diagnose ASDs?
A: The diagnosis of autism is a clinical diagnosis, it’s not a pathological diagnosis.
Q: So your testimony is that pathology is not something that would be used to diagnose autism, is that correct?
A: That’s correct.
Q: You were also asked about how many total brains have been used in this long series of neuropathological studies involving autism. How many?
A: Our own series is 9.
Q: So you had 9 brains?
A: 9 that we examined.
Q: And that generated the series of papers?
A: Yes. Hutzler’s is 8 and Bailey’s is 6. And those are the series. There are individual case reports as well.
Q: And the first series was published in 1994, is that the first one?
A: No, the first case was ours in 1985.
Q: So in the series of papers you’re talking about, there are 23 individual brains represented by multiple authors?
Q: So whatever findings you extrapolate would then be used to apply across the entire spectrum of the population of people with autism?
A: Well, they’re randomly collected; there was no special reason to pick these, other than that these were people with autism.
Q: Do you have an idea from the time you started doing this to the present, how many people with autism there are in the US?
A: Current CDC numbers are 1 in 156.
Q: So in a population of 300 million, if 1 out of 156 has autism, we would be talking about several millions of people with autism?
Q: So these 23 brains are a sample out of many millions of people; is that a fair statement?
Q: Is there any way to get a neuropathology of a living person with autism?
A: Yes, MRI scans would be one way to do that.
Q: What other ways?
A: Well, I suppose there are ways to look at organization with a functional MRI, there are PET scans.
Q: Aside from autopsy, is there any way that one can take an autistic individual and learn the neuropathology in that individual?
A: The resolution of these other techniques is such that it’s very unlikely that you’d pick the neuropathology up.
Q: So it’s unlikely that imaging would pick them up?
A: There’s some, but nothing that would pick up the detail that I showed.
Q: This is detail at the cellular level?
Q: For this, you would actually need tissue autopsy?
Q: So imaging wouldn’t establish it?
A: They tried, but it’s not very good.
Q: So short of autopsy, you can’t really look at the brain of an autistic person to figure out what pathology is going on?
Q: Except with brain size. Now, with brain size, that would be measured by head circumference, is that correct?
A: It depends on how old you are.
Q: How does it depend?
A: I’m not sure where the age cut off is. In young children, you can measure with a tape, and there’s a very good correlation between brain size and head circumference. Later on in development, it becomes unreliable.
Q: And when you say it becomes unreliable, you mean head circumference no longer is a good surrogate for measuring brain volume?
A: It’s very accurate early on.
Q: As a person gets older?
A: It’s less accurate.
Q: Now there have been a series of studies done that looked at head circumference and head growth patterns among children?
A: Normal or autistic?
Q: Autistic. Those are the ones cited in your report, and there’s an effort to describe trajectory of brain growth?
Q: My recollection of Hazlet, which you cited, is that the brain growth among autistics and controls was roughly the same for the first year, but then widely diverges. Is it your recollection?
A: I don’t recall. What did I say in my report?
Q: What you say in your report on p. 4, last full paragraph, is that several studies have “reported head circumference of autistic individuals began to diverge at 12 months of age.” So did I capture that accurately?
Q: Now look at the Dawson study that you cited. What Dawson says is that the differences in head circumference were confined to the first 12 months?
Q: So you’ve got one study that says head growth is the same and then diverges at 12 months, and the Dawson study says the differences are confined to the first 12 months.
A: That’s right.
Q: Now let’s look at the Courchesne 2003 paper you discussed in your slides. There, he sees no statistically significant increase at 3-5 months, but a statistically significant increase at 14-16 months. Courchesne shows different results from Hazlet or Dawson, is that correct?
Q: Now Dementieva describes a “sudden and excessive increase at 1-2 months.”
Q: So you have these four studies that are divergent and at times conflicting, correct?
Q: So the diversity in these head studies is further discussed in Dr. Layner’s paper, a government exhibit. Susan Fallstein [presumably an author on the study], is she a colleague of yours?
A: I know her.
Q: Does she work at Boston University?
A: No, she was at Tufts and now she’s back in Florida.
Q: In that study, it states that the implications are that there is a very wide distribution of head circumferences among people with autism?
Q: And they say that the diversity of head circumference data potentially reflects the diversity of clinical presentation of symptoms of autism, correct?
A: I don’t remember correlations with clinical features off the top of my head.
Q: The authors highlight that the increased variance of head size distribution underscores the clinical heterogeneity of autism; is that your recollection of what the study stands for?
A: Yes, one of their conclusions.
Q: It talks in the same paragraph about studying the dimensional features of autism.
What are the dimensional features of autism. Would that include time of onset of autism?
A: That I don’t know.
Q: Would time and progress of disease be important considerations in studying that disease?
A: Yes, of course.
Q: Is there anything in this paper to suggest that data was collected specifically on children with regressive autism?
Q: In any of the studies, aside from Vargas, were data collected only on people with regressive autism?
Q: In the 23 brains for which you described neuropathological findings, did any of those brains have specific information like that, about whether the individuals suffered from regressive autism?
Q: Was there no information about severity of symptoms of autism?
A: There was information on severity.
Q: What information on severity?
A: Bailey’s cases are much more severe than the others.
Q: In Bailey or anybody else, is there any information about age of onset of the symptoms?
A: I have not looked at those papers for that issue.
Q: I want to turn to your slide presentation, on p. 4, the template of developmental events. There’s a very large scale for events that happen in the 40 weeks of gestation from conception to birth. Then after birth, the only thing listed is a compression of the first 4 years of life. You don’t mean to imply that the only thing that happens is abnormal brain growth after birth?
A: Certainly not.
Q: After a child is born, there’s a significant amount of brain development?
Q: Synapses are being formed?
Q: Synapses are being pruned so they’re more functional?
Q: Axons and neurons are migrating through different regions of the brain?
A: Yes and some are being eliminated.
Q: Organization of cells is going on?
A: Mainly organization at that stage.
Q: This is a process that goes on at least through the first 2 years of life, correct?
A: Way beyond that.
Q: And a lot of that process is mediated by glial cells?
A: Well, they’re part of it, because they support the neurons, but the major thing is neurons.
Q: But the movement of neurons and the organization of neurons, cells such as oligodendricytes and astrocytes, play a significant role?
A: In later brain development, from birth to 2, there’s a lot of brain development there.
Q: Is that with the myelin sheathing?
A: Yes, there’s an explosion of that.
Q: A lot of that activity relies on properly functioning glial cells and a glial network, correct?
Q: And anything that interferes with the normal functioning of glial cells could interfere with neuronal end points?
A: That’s a reasonable statement.
Q: On p. 6 of your slides, this is a section of brain, a thin slice that you put on a slide?
Q: Is it stained?
Q: What is it stained with?
A: It stains nerve cell bodies and glial cell nuclei.
Q: Can you see any glial cell nuclei?
A: No, they’re too small.
Q: Is that a function of the stain or the resolution?
A: It’s certainly a function of the size. Glial cells you have to identify under very high power.
Q: What are criteria for identifying glial cells?
A: The gold standard is immunostains, called GFAP.
Q: And that’s the gold standard because it’s immune reactive?
A: Only with astroglial cells.
Q: And it doesn’t stain the neurons because it wouldn’t have the immune cells on the surface that would attract the stains?
A: Yes, and it wouldn’t stain other species of glial cells either.
Q: On p. 8, is this your slide?
A: Yes, that’s ours.
Q: Was GFAP used in any of these?
A: No. Our own brains are embedded in a substance that doesn’t allow for immunostains.
Q: So none of your slides are stained with GFAP and are not at a resolution that would allow for identification of astroglial cells?
Q: There was a lot of talk of neurons, but there was no discussion of glial cells in the imaging?
A: Pretty much what’s known about glial cells is in that Vargas paper. Our papers have said very little about them.
Q: And you didn’t say very much about them, because of the basic work up of tissue; you couldn’t have done the work, right?
Q: On p. 9, these are the Purkinje cells. You said something about GABA-ergic. What’s a GABA-ergic neuron?
A: It’s an inhibitory neuron; that’s the inhibitory transmitter in the brain.
Q: Are Purkinje cells GABA-ergic?
A: Yes they are.
Q: So Purkinje cells would excrete the neurotransmitter GABA which is an inhibitor?
Q: So it’s the main inhibitory neurotransmitter in the brain?
Q: It’s the flip side of the glutamate coin?
Q: So if there was excess glutamate in the brain, GABA-ergic cells, like Purkinje cells, would be secreting GABA to maintain homeostasis? Is that correct?
A: You have to rephrase that.
Q: If there was excess glutamate in the brain for any reason, feedback from Purkinje cells would be to release GABA to inhibit and bring back balance?
A: I don’t know if particularly Purkinje cells would do that, but the brain is filled with GABAergic neurons; it wouldn’t have to be that one.
Q: Are there other cells that are not neurons that are GABA-ergic?
A: I don’t know of anything other than neurons [that are GABAergic].
Q: Are there any astrocytes that are GABAergic?
A: That’s not in my expertise, but I don’t think so.
Q: On p. 12, the inferior olive in childhood and p. 13; are these images from your lab?
Q: And p. 16 and 17; are these images from your lab?
Q: From the same 9 brains?
A: No. We have a lot of studies going on in our lab about autism, and I do the neuropathology on some of these studies. These are frozen samples from other brains.
Q: And these are not published anywhere?
A: Right, these are just nice examples. Cells in the white matter are shown by several other people, it wasn’t just me.
Q: In the slides that you were just looking at, unpublished case control ones, in the staining on these, did you do analysis to identify glial cells?
A: No, these were all done for neurotransmitters.
Q: Which particular neurotransmitters?
A: Mainly GABAergic system.
Q: Was there any work involving glutamate?
A: Before I left to come here, I checked, and there is nothing ready to publish on glutamate….
Q: In any work in your lab, published or unpublished, have you ever looked for inflammation in the brains of autistic people?
Q: Let’s talk about the Vargas and Pardo work. There’s been discussion of a letter. This is personal correspondence between Dr. Pardo and yourself?
A: I presume so.
Q: It’s dated May 13, 2008, but it addresses a conversation that you had last year?
A: That’s correct.
Q: In reading Dr. Vargas’s paper, p. 13, what the authors say is “These observations do not support the previously proposed hypothesis that changes in the cerebellum in autism result solely from developmental abnormalities, olivary cerebellar circuits and a reduced number of Purkinje cells.” I was just reading that, but is that correct?
Q: [And is the citation for this statement to your paper?]
Q: So when this paper came out, the authors were saying, our results are not consistent with the work Dr. Kemper and Dr. Bowman are engaged in?
Q: Was it after this paper came out that you had a conversation?
Q: How did conversation come about?
A: We were at a meeting together. He had a wonderful paper about neuroinflammatory response. I was just curious about what he thought about all the developmental problems that we found and how he felt it fit in.
Q: So it was more of a general conversation of two scientists talking about work overlap?
[Asks a series of questions about quotations from the Vargas paper on neuroinflammation; at 37 minutes on file 4 of 5/22]
Q: The argument has been made that there are protective pro-inflammatory cytokines released, and you were relying on this paper for support of that proposition?
Q: The authors of the paper note that TGF-1, while it is anti-inflammatory, it is actually released by cells that are dying?
A: That I don’ know.
Q: Well, it says it in the paper?
A: Yes, OK.
Q: Your understanding is that while you have this inflammatory cytokine response going on, while it might be protective to some cells, it is evidence that adjoining cells have recently died?
A: I don’t know.
Q: On p. 14 in the Vargas paper….the authors conclude that neuroglial reactions… neuroglial refers to microglial and astroglial, correct?
Q: So neuroglial reaction involves microglia, the ‘macrophages of the brain’? So in the mechanism they describe, the cerebellum is the center of common and chronic inflammation in autistic patients?
Q: They don’t describe a particular cause?
Q: They don’t limit it to earlier neuropathological conditions of these patients?
A: It’s one of the possibilities.
Q: So it’s one possibility that neuroinflammation is related to the pathology that exists that is prenatal?
A: Well, I’m not sure if it’s “is” or “can”.
Q: Well, that’s what I’m trying to get at; it “can” be, but it’s not necessarily so?
A: As I told you, I’m not an expert on this stuff. What I know is what’s in Dr. Pardo’s letter and here.
Q: I understand, but you were asked questions as to what your understanding and your opinion was. So your opinion is that one of the possible explanations for neuroinflammatory process is pathology that has its origins prenatally?
A: I would say likely.
Q: It is also possible that this process reflects events that happened postnatally?
A: I suppose so.
Q: It’s also possible that these neuroinflammatory processes reflect response to a toxin or other environmental, exogenous factor that could have triggered a neuroinflammatory process?
A: Well, they actually state that they don’t think that it’s due to an exogenous toxin.
Q: Where does it state that?
A: [Dr. Kemper looks in the article; long silence] I don’t see it in here.
Q: So there’s nothing in the peer-reviewed papers that rules out toxins as causing this neuroinflammatory process?
A: I would have to read it again with that mind.
Q: You just read Dr. Pardo’s letter, and there’s nothing in there ruling it out?
A: I didn’t see it.
Q: Let’s look at Dr. Pardo’s conclusions. Dr. Pardo is hypothesizing that environmental factors, in the presence of genetic susceptibility and the immunogenetic background of the host, influence the development of the abnormalities they observed.
Q: And it also attributes this hypothesis to neuroinflammatory changes responsible for the generation of autistic symptoms?
Q: So not only does Dr. Pardo’s letter not rule it out, it is advanced as a hypothesis in the published literature.
A: Let me read it again. It refers to neuronal circuitry.
Q: And neuroinflammatory changes?
Q: They hypothesize that environmental factors could play a role?
Q: If you look at the chart, Figure 4 of this paper, there’s a chart that includes environmental factors in the development of the autistic phenotype. Under environment, toxins are specifically noted, and under phenotype, regression is specifically listed, correct?
Q: So this paper does postulate a hypothetical role for environmental exposures in creating neuroinflammation that ultimately lead to the regressive phenotype of autism?
A: That can be an interpretation.
No further questions.
Re-direct of Dr. Kemper
Q: You were asked some questions about brain development during first 2 years of life? A: Yes.
Q: And you were asked about role of microglia and astrocytes and the development of the brain?
Q: Do you have any evidence that thimerosal affects that process of brain development? A: I’m not aware of any evidence.
Q: Is the process that you’re being asked questions about involving microglia and astrocytes, is that what Dr. Kinsbourne is saying is causing autism as a result of exposure to thimerosal?
A: Not to my understanding.
Q: You were asked about your own work and the brains that you looked at. Can you describe how you look at slides and what that process entails? The process from when you get a brain sample and how you go looking at these samples.
A: If it’s an autopsy, we prepare the section with whatever stains are necessary, we look at them, describe it and draw a conclusion.
Q: How much of the brain do you look at?
A: 4 or 5 blocks.
Q: How long does it take you to look at single section of the brain?
A: 10-15 minutes.
Q: Is this a meticulous process?
A: It depends on the nature of the brain.
Q: You were asked about personally looking for neuroinflammation. And you answered that you didn’t look?
A: Yes, we didn’t look.
Q: But the Vargas group did look?
Q: And they found astrocyte activity was increased?
Q: And that’s not consistent with Dr. Kinsbourne’s hypothesis?
Q: You were asked about the Vargas research and Dr. Pardo’s letter. Is Dr. Pardo the senior researcher on that study?
A: That’s my understanding.
Q: Did he say to you that he didn’t think it was a neurotoxin that was causing the neuroinflammation?
A: I know I did hear it from him.
Q: Please look at Dr. Pardo’s letter. It states that “the staining is inconsistent with toxic material…”
Q: Is this what you meant when you said you read it somewhere before?
Q: So you think that’s what you were referring to?
Q: During your conversation, did you also discuss whether neuroinflammation could be consistent with a response to abnormal development?
Q: And is that opinion consistent with the letter that he sent?
Q: And in your opinion, and based on your 30 years of experience working with autism, do you believe that it is more likely than not that neuroinflammation is a response to developmental abnormalities that occur prenatally?
Re-cross of Dr. Kemper
Q: In Dr. Pardo’s letter, the final letter doesn’t say that astrocytes were unaffected by pro-inflammatory agents released by microglia?
Q: There’s no mention of astrocytes being unaffected by toxic oxygen species?
A: No, there’s no mention at all.
[specific questions about the process of neuroinflammation; at 55 min. in file 4; Dr. Kemper says he can’t answer]
Q: As a scientist, would you rely most on peer-reviewed literature?
Q: And if it’s in a really good journal by a really good author, it carries even more weight?
Q: And if you were involved in an investigation, and had to rely on an unpublished, personal communication from one scientist to another? Which do you think would be more reliable?
A: I guess it would depend on my opinion of the scientist.
Q: We’re talking about a disinterested observer deciding scientific facts. What would rely on more -- personal correspondence or peer-reviewed literature?
A: The latter.
Q: No further questions.
Mary Holland teaches at New York University Law School and may be contacted HERE. Thank you to A-CHAMP for providing this important service. Thank you to A-CHAMP for providing this important service. A-CHAMP, Advocates for Children’s Health Affected by Mercury Poisoning, is a national, non-partisan political action organization formed by parents in support of children with neurodevelopmental and communication disorders.