Albrecht Fröhlich

Transcription

>> I'd like to introduce Stuart Hameroff from the University of Arizona. He is going to talk about Clarifying the Tubulin bit/qubit, and then five minutes on Defending the Penrose-Hameroff Orch OR model. >> HAMEROFF: Thank you very much. Thank you all for being here. Thank you Hartmut and Google for hosting us and you all for coming. The title of talk is Clarifying the Tubulin bit/qubit, Defending the Penrose-Hameroff Orch OR model of quantum computation in microtubules. Jack--Travis Craddock and Jack Tuszynski are my co-authors, and Jack is here. Penrose-Hameroff Orch OR orchestrated objective reduction is a theory of consciousness, based on quantum micro--quantum computations in microtubules. OR is Penrose's mechanism for self-collapse of the wave function, Orch is orchestration which is my biological contribution, how this could happen, and both be isolated and interact with the environment. But I'm not going to talk about that specifically. First, I'm going to give you a little bit of an overview about microtubules. We've heard quite a bit about some of the technical aspects, but I want to kind of backup and give you the big picture a little bit. Consider a single cell paramecium which can swim around, avoid obstacles, avoid predators, find food, learn; if you suck it into a capillary tube, it escapes. If you do it again, it escapes more quickly. It can find a mate and it can have sex. This is actually a picture of two paramecium having sex. It's probably the only X-rated picture we'll see today. And it doesn't have any synapses. It doesn't have any nervous system per se. How does it do it? It uses structures called microtubules. And I often kid my AI friends; they should be worry about simulating a paramecium before worrying about a brain. So microtubules seem to be the nervous system of the cell. And here's a beautiful immunofluorescence micrograph of a double-nucleated cell. The yellow is the microtubules. The red is the actin. And the microtubules form the kind of cytoskeletal architecture and bone-like support, but we also think they're the cells onboard the computer. In cell division, microtubules are in the red, the blue--the chromosomes are in the blue, and the yellow are centrioles which anchor the mitotic spindles and microtubules and organize mitosis. And if this goes wrong, you can get cancer because you get an abnormal genotype. And the mechanism of this is still poorly understood. The centrioles are made up of nine triplets of microtubules so each of these would be a microtubule or part of a microtubule and you get nine of these in forms of mega cylinder. And this is the same structure also found in cilia, the appendages that come out at the end of the paramecium. And we also have them in our body and they're found in the retina of the eye and they're found all over the place. And in cell division, centrioles have two of these barrels in this odd, perpendicular arrangement. They each replicate and separate and the mitotic spindles anchored to them and pull apart the chromosomes. It's interesting there are also optical--optically-sensitive--Albrecht-Buehler, who was mentioned earlier, showing that these centrioles detect photons and orient the cell in a particular direction towards the light. And it's interesting that their shape and their geometry, their dimensions are--could accommodate photons and they could be some kind of quantum optical devices or at least optical resonators of some sort. I just want to mention that. I'm not going to talk about it per se. So inside neurons, microtubules and dendrites are found in this interrupted type of network, intercorrected--interconnected by microtubule-associated proteins. So they form a sort of network and we think a computational network inside cells, inside neurons of a neural network. It's kind of a fractal subdimension inside neurons. Now if you're microtubules go bad, you get Alzheimer's Disease. There's actually two lesions in Alzheimer's; amyloid plaques, which get--for some reason get most of the money in research. But the real damage is done by neurofibrillary tangles inside the neurons which are due to tau proteins which normally--here, it says stabilize microtubules but they do other things as well. When they get all screwed up, the microtubule disintegrates, that the tau gets hyperphosphorylated and the microtubule disintegrates literally. You get the kind of crinkly neuron and you lose cognitive memory and eventually consciousness. So Alzheimer's is a disease of microtubules. Here's a close-up of a disintegrating microtubule. Now, the tau protein which is thought previously just kind of hold the microtubule together, turns out that it does something more. It acts as a kind of a traffic signal for motor proteins which transports synaptic materials. So if a synapse downstream in a dendrite, say, needs a particular enzyme or precursor or receptor, it's often synthesized more proximately and transported by these motor proteins which carried along as cargo, and they often have to jump tracks and switch microtubules in the branching dendrite. And it's been a mystery how they seem to know where to go. It turns out that the tau, at specific locations on the microtubule are kind of traffic signals and tell particular proteins, like this guy right here, where to jump off. And so their placement is critical. Now how do they know where to--where to be? Is there something--are they that smart or is there some kind of encoding in the microtubule itself? So I'm going to briefly talk about--a little about memory, some recent evidence about memory that Travis Craddock, Jack Tuszynski and I have done. In long-term potentiation, calcium comes in through the cell and activates CaMKII, calcium/calmodulin kinase II, which then rapidly--in learning, rapidly distributes throughout dendrites and actually throughout many parts of the neuron or group of neurons quite rapidly and they're associated with microtubules. And somehow--and they phosphorylate something in the cell which stores memory. Now, synapses are created with memory, but synapse--the proteins of synapses are very transient. It last hours to days, but memories can last a lifetime. So the question is where memories may be stored. And the CaMKII is a very interesting molecule. It's snowflake-shaped holoenzyme which when activated and phosphorylated by calcium coming in, transforms into this sort of insect-like nano-poodle, we like to call it, with six legs extending up and six legs extending down, each of which can phosphorylate a substrate. So the question is what's the target for phosphorylation which memory can be stored? Well being microtubule fanatics, we thought it might--has something to do with microtubules. And so here's a microtubule with two different scales; a microtubule then a close-up of the tubulin subunit showing the phosphorylation sites on tubulin with the C-terminal tails and--so the question is, how did the CaMKII relate to this microtubule? And it turns out it matches perfectly. Here's the CaMKII here with the top set of legs taken off and overlying the A lattice and B lattice. And [INDISTINCT] talked about the different lattices. You can see here that they matched up perfectly. So each of these legs can either phosphorylate a substrate or not, therefore, it conveys a bit of information. And six bits on a--an ordered array of bits is a bite. So basically, the idea is that each of these is--can convey a bite of information onto the--onto the tubulin lattice. And sure enough with a little flexibility in these extenders, the nano-poodle can deposit information on the microtubule and perhaps move--march along it by hydrophobic interactions and can convey synaptic information to the microtubule and, here, mechanisms at the level of individual amino acids where the phosphorylation can occur either through the C-terminal tails or by a different mechanism. So basically, we show the possibility for memory storage in microtubules via calcium-induced phosphorylation of CaMKII which then deposits the information for long-term--longer term storage on microtubules. And depending on a couple of factors; A lattice versus B lattice, and whether alpha and beta can be--can be phosphorylated, the number of bits per bite for CaMKII, in this case, it's two to the sixth or 64. In this case, it's about 429. In this case, it's 5,281 just from one CaMKII. So the amount of information capacity in the--in the--conveyed through synaptic interactions, CaMKII on microtubules, is enormous. And the number of--the amount of information--storage capacity in a microtubule even in one neuron is enormous. So that suggest that microtubules are good information processors and many of us have thought of this for a very long time. And on the left, you see a picture of a microtubule with its sub-unit proteins tubulin which can switch between two states and possibly also superposition on both states, the quantum bit or qubit. I got interested in microtubules in 1972 when I was in medical school and worked for about 20 years on the idea that they process information strictly classically and developed the molecular [INDISTINCT] models with a number of physicists, I meet Jack along the way, Steen Rasmussen of Los Alamos and others, modeling them as sort of the game of life. This cellular automata shown here are based on dipolar interactions between the tubulin and show that, with some basic assumption using Frohlich oscillations as a clocking mechanism that microtubules could process information in principle, based on a few assumptions, were very efficient computational devices for both storage and processing of information. In here, it just shows a sequence. Initially, we thought the clocking sequence was Frohlich gigahertz, but the recent information we're perfectly happy with eight megahertz clocking that Jirí Pokorny has discovered and [INDISTINCT] been talking about. So each of these steps could be--could be occurring at eight megahertz and by a--sort of a cellular or molecular automata type of mechanism, but so far this is strictly classical. Now in the late '80s, somebody said to me, "Okay, let's say you're right, how would that explain consciousness?" And I had to admit that they were right. I didn't really know. Fortunately, I read Roger Penrose's book, 'The Emperor's New Mind', where he had a mechanism for consciousness based on quantum computation, but he didn't have a good qubit in the brain. So I suggested microtubules were his quantum computers and tubulin with the quantum bits and he agreed. He liked the geometry of the microtubules and we developed the model in the--in the early mid '90s and it was almost immediately attacked by a number of people for various reasons. It--it's threatening to many people for different reasons. So the basic idea is that in a microtubule, each tubulin can be a quant--can be a qubit so it can be in two states in the superposition of both states. And consciousness occurs because there's a superposition that reaches threshold for the Penrose objective reduction and consciousness occurs here, involving you know, hundreds of billions of these. So I'm not going to really talk about consciousness per se, but just about fundamentals of how individual tubulins can switch. And we also suggested that quantum states can tunnel between neurons through gap junctions which, it turns out the gap junctions mediate gamma synchrony which is a neuronal correlate of consciousness. So we were attacked, as I said, almost immediately and--by philosophers, and physicists and so forth. Nothing too serious. And then in 2000, Tegmark--Max Tegmark came--tried to prove the obvious. The brain is too warm and wet for delicate quantum effects and had this equation for the decoherence time and published it in Phys Rev E. And then a year later, Jack, and Scott Hagan and I published--used his own decoherence formula, corrected for some of his errors in terms of how he characterized our model. In other words--for one example, he had superposition of a soliton separated from itself by 24 nanometers where our superposition separation was the fermi length, the diameter of an atomic nucleus. So that alone--where as--so he calculated decoherence time of 10 to minus 13 seconds, that alone saved us seven orders of magnitude. A couple of other corrections we got down into reasonable for the Orch OR model of hundreds of milliseconds. Again, this is all theory, but we use his--it countered his theory and more recent findings of quantum coherence in Biology is supporting this at least in a general way. Okay. So more recently, a group of Australian biophysicist, physicists, chemists, actually a very impressive group if you look at their credentials, published a series, at least two so far or maybe more, attacking us. And in two papers, one in Phys Rev E and the other one in PNAS; pretty good journals. And the first one, the one I'll address first, McKemmish et al. The title was "Penrose-Hameroff orchestrated objective-reduction proposal for human consciousness is not biologically feasible." Well, that's pretty harsh. And they also said in the abstract that not only is it--is it not feasible, it couldn't be fixed by any conceivable modification. So I don't know if they're psychic or just, you know, dancing on our grave or what. The other paper, which actually came first, "Weak, strong, and coherent regimes of Frohlich condensation and their applications to terahertz medicine and quantum consciousness," challenges also on the basis of the Frohlich condensation that Jirí Pokorny was talking about. So, objection number one--I'm going to take the McKemmish et al paper first. They start up by saying--which is true--the Orch OR asserts that tubulin states are regulated by London force dipoles, the van der Waals-London force dipoles, in hydrophobic pockets within each tubulin. So I want to take a minute and explain about hydrophobic pockets. Luke was talking about non-polar regions in the GABA receptor, and Elizabeth was talking about London forces or van der Waals forces. So when a protein folds, here's a simple protein. And the non-polar groups like [INDISTINCT] escaping water coalesce and they don't have to be all these aromatic groups. It can be any non-polar group forming a hydrophobic pocket shielded from water. So when you--when they say the brain is too warm, wet, and noisy, it's not really wet in the hydrophobic pockets because water is excluded. So, hydrophobic pockets are water excluding regions inside proteins consisting of non-polar--of groupa such as aromatic amino acids, phenylalanine, and tryptophan and so forth. And if you--if you think about this non-polar solubility region, it's a very small portion of the body. Actually, this--you can think of this triangle as a body ground up and then distributed in terms of solubility spaces, different solubility spaces, and you can see where different solvents will dissolve. And as an anesthesiologist, I can tell you that anesthetics binds in this little region here, aromatics, right down here; very non-polar. So this is where consciousness is, okay? You can--the rest of it doesn't really matter in terms of consciousness. Well, supportive of course. But this small region of this diagram is a non-polar region where anesthetics act and where consciousness resides, but I'm not going to talk about consciousness. So--well, I guess I did. Okay, so getting back to the McKemmish et al attack. They said we assert, which we do, that tubulin states regulated by London force dipoles in hydrophobic pockets within each tubulin. And then attempting to invalidate our research, and they said that the electron cloud of a single benzene it is--suggesting it to be analogous to, and indicative of, hydrophobic pockets, cannot be a switch because it's completely delocalized. So there's two basic--two ways to look at a benzene which is basically a phenyl ring on a phenylalanine, Valence theory that there's a resonance between two different states for the--where the double bonds flip back and forth between two possible locations or you can just say it's completely delocalized in molecular orbital theory. They said, "Well, benzene is, in molecular orbital theory, there's no switching because it's completely delocalized therefore the hydrophobic pockets can't be switches." Well, we agreed that a single benzene could not be a switch because it takes two to tango. And in our case we--London forces, if you have two benzenes together they're going induce induced dipoles. These are inducible--induce induced dipoles. So, two clouds come together. The electrons in one repel the electrons in the other and then they oscillate back and forth. So these are the double bonds flipping back and forth and this is kind of a new schematic for it. And if you have four, maybe it will look something like this. And Hartmut borrowed this for the--for the logo, for this workshop, but put the Google colors in which I'm quite proud of actually. So this is what van der Waals-London forces are. This is shown for a neon atom, but it works for any neutral--electrically neutral non-polar, two electrically neutral non-polar groups where the blue dots are the electrons and the electrons in one repel the electrons in the other and this draws in together the van der Waals-London attractive force. So they get closer. They actually repel a 100 times stronger. So this very weak but important force mediates, for example, how anesthesia works to erase consciousness in that little piece of the phase diagram. A hundred--over 100 years ago, Meyer and Overton showed that the potency of any anesthetic gas correlates with its van der Waals--well later, it was appreciated, it was van der Waals-London force binding in a non-polar medium akin to olive oil over wide ranges of molecular structure and concentration. So getting back to the tubulin bits. So our--one of our early models show the tubulin switching between two states due to a hydrophobic pocket where there's two aromatic rings, where we only show a piece of each with the electrons kind of flipping back and forth. And here's kind of a more recent model and here's the more current model. But maybe McKemmish et al didn't realize we were talking about two different rings because they said we only had one phenyl or benzene ring in there. Actually, there's many phenyl and indole rings in a tubulin. This is some recent work from Travis Craddock mapping the phenyl--phenylalanine with the aromatic rings, and indole rings, and many of these are within three ångström of each other or less than four ångström, 3.7 is where the van der Waals--van der Waals forces balance out. So many of these are close enough to form coalescences of these aromatic groups to get hydrophobic pockets with a lot of electron resonance and mobility--electron mobility which Luke was talking about. And with a little imagination, you can actually see pathways for topological qubits that [INDISTINCT] was talking about. We'll get to that later. And these aromatic groups can have a particular orientations for their London forces. And I think maybe Jack is going to talk about this so I'm going to skip over it for now. So conclusion to objection number one; McKemmish et al missed the boat entirely. The hydrophobic pockets include minimally two non-polar groups which utilize van der Waals London forces. It takes two or more to tango. So they said while one benzene ring wouldn't do the trick for you, we agree completely. That's--we never said that, we're talking about at least two and actually coalescences of multiple non-polar group of electron resonance capabilities. Okay, another aspect of the decoherence as well into this--we've heard about quantum coherence in photosynthesis and some in microtubules. A paper--in 2003 by Ouyang and Awschalom connected quantum dots by benzene rings, okay? So these are single benzene rings, but they're just looking at quantum spin transfer and they found a very efficient quantum spin transfer connecting the quantum dots through these fennel rings basically, benzene rings, the same--exact same rings found inside the tubulin and other proteins. And as far as temperature, they're starting at absolute zero, when they got to about minus 70, there was a huge jump up in efficiency which persisted out to bring temperature by connecting it through this fennel or benzene ring. This is quantum spin transfer, not exactly the same as we're talking about. But it shows that a quantum effect can be enhanced by temperature and that this fennel or benzene ring can be utilized in this way. The same sort of structure--and I think Hartmut and Luke alluded to this, are found in psychoactive drugs including powerful psychedelics, in fact any psycho-active drug including chocolate. LSD, DMT all have these non-polar indole-like rings here. And in the 1970s, we're shown that the--looking at a series of psychoactive--psychedelic drugs actually, hallucinogenic drugs, that their potency was related to their ability to donate electron resonance energy from the drug to the receptor. So the potency of the psychedelic is related to its ability to donate electron resonance energy, quantum electron energy to the receptor, to the system and that seems to push the person into an altered state. Okay, so that was the first--the first objection. The second objection was this. It concern GTP hydrolysis and tubulin switching. McKemmish et al say, "Experimentally, the only identified aspect of the conformation of the tubulin heterodimer inside a microtubule depends on whether GTP or GDP is bound to the beta subunit and is the only process to be ascribed to the conformational change depicted in the Orch OR proposal." Well, Jirí talked about how the Frohlich condensation, the Frohlich vibrations are fueled or pumped in a cell by being near mitochondria taking advantage of mitochondrial energy, but not GTP or ATP directly. Just being in the field is enough to drive the coherence. Now, the other point is this conformational change. And it's true that in our--in our cartoons, and these are cartoons admittedly, that there's a significant conformational change. The tubulin is switching states by, you know, at least 10% of its volume. Now, we did that to indicate a change of state, but actually in reflection, we don't really need a conformational change to register information. It can be strictly a dipole state. And although we showed these cartoons as having conformational switching, we never really utilized the conformational change in our model except that for the quantum state, the superposition separation is of the diameter of a carbon nucleus, the fermi length, 10 to the minus 8 centimeters is very, very small. So that gives you the superposition separation which gets in the quantum gravity stuff. But as far as changing the state for a computer, we don't need a conformational state. We're perfectly happy with an electronic state that can be read out by its neighbor. That's all we need. Now, the GTP business is also a mistake on their part because when GTP hydrolyzes on the--from a GTP to GDP, it does cause a conformational change so that could be how they got that idea. It releases the phosphate bond, but it's irreversible. Once this happen, the microtubule starts to depolymerize. It may depolymerize at one end and then repolymerize in the other which gives rise to treadmilling or may fall apart at both ends which gives catastrophe or dynamic instability which is why they rapidly shrink and grow, shrink and grow. But we point--we looked primarily for Orch OR in dendrites, in dendritic microtubules where the microtubules are capped and don't undergo GTP hydrolysis, don't undergo treadmilling nor depolymerization-polymerization cycles. They're quite stable which is why they would be good for storing information in dendrites. So they're wrong on this point also. So the conformational switching greater than one fermi length, the diameter of one carbon nucleus, is not required. Electron dipole state switching is sufficient. So this is kind of a new model of--for the tubulin bit or qubit oscillating at eight megahertz as Jirí was talking about and [INDISTINCT], between two states. You'll notice there's no conformational change here. If you didn't see the electron or couldn't measure the dipoles, it would look exactly the same just like if you looked at a computer motherboard, you might not be able to see all the information going around because you can't see the electrons. But we actually have--now have multiple hydrophobic pockets each consists--each containing a number of non-polar groups and aromatic rings and this is perfectly fine for Orch OR. The only conformational state change we need is down at the level of--for the qubit function, is down at the level of the atomic nuclei. So objection two; conformational switching is not required, dipole state switching sufficient, Orch OR applies to dendritic microtubules which are capped to prevent GTP hydrolysis, treadmilling and dynamic instability. Orch OR does not require GTP hydrolysis so they're wrong on that point also. Now, I want to talk a little bit about topological qubits. [INDISTINCT] gave some very elegant results and I'm not sure my understanding is up to, you know, from the quantum computing point of view. So I'll just tell you my interpretation of topological qubits which may not match--which may not be right or match what he said. But historically, how this got started for microtubules was that Roger Penrose invited me and Jack Tuszynski to the 1998 meeting at the Royal Society in London of quantum information to talk about microtubules. And actually, Jack was the--was the skeptic and I was the proponent and since he's kind of come over our side, thank you. But afterwards, there was a--John Preskill gave a talk on topological quantum error correction and topological qubits and relating to [INDISTINCT] idea and so forth. And as I understand, that the idea is that if you have some kind of lattice and the information can flow through one pathway or another pathway then the pathway itself can be the bit. So in the case of microtubules, they have this Fibonacci geometry, so you can follow this pathway, or you can follow this pathway, or this pathway or this pathway, or a number of other pathways. And if each tubulin was a bit--if each one of these guys was a superposition and represented a quantum bit or qubit then it would be susceptible to decoherence. Each one--decoherence of any one subunit would mess up the whole quantum computation. However, if you make the pathway, the bit or the qubit, then if one of these guys gets knocked out or whacked, it's going to get pulled back in by its neighbors and so it becomes resistant to decoherence. That's the whole point about topological qubits and also topological quantum error correction which is a very similar idea. And these spiral pathways follow the Fibonacci geometry; 3, 5, 8, 13, 21 and this was one of the things that got Roger very interested in microtubules because he is a geometer at heart and that's one of the things that captivated him. And [INDISTINCT] is now shown something like this, in microtubules, that it actually occur. So it's also--it might be something like the Aharonov-Bohm effect where you have the opposite pathways or alternative pathways, each of which can be a bit or a qubit. And if you look at the multiple hydro--non-polar hydrophobic pockets and line them up, with a little imagination, you could see how their--states of each one of them could influence conductivity of ballistic conductance or quantum coherence moving through the lattice around--along particular helical pathways in the microtubule. Okay, objection number three. This was from the earlier paper from the same group, same author group, Reimers et al, about three types of Frohlich condensation: weak, strong, and coherent. And Jirí talked about this a little bit. Based on--and they based their conclusion on a simulation of a linear chain of tubulins to represent three dimensional microtubules. They concluded that only weak Frohlich condensation at eight megahertz, and they cited Pokorny--so they essentially validated Jirí's work. And since they were so negative and skeptic about everything else, I think it was kind of remarkable that they--that they, you know, they agreed to that and they liked it. But they said it was only weak Frohlich condensation and that Orch OR requires strong or coherent condensations. For example, Bose Eins--something like Bose Einstein condensation. Well, it seems to me that we don't want Bose Einstein condensation because in that case all of the--all of the subunits would be in the exact same state and you couldn't really do any computing that way. You need--you need the--you need some--the possibility for each of the subunits to be in a different possible state. And so we think that weak is perfectly fine for us. Orch OR does not require strong or coherent Bose Einstein condensate, requiring only synchrony and entanglement. So we just wanted the Frohlich condensation, Frohlich excitations to synchronize the operations like in a cellular automata. In a cellular automata, you need--you need a clocking frequency. And a weak condensation is fine as long as you can have entanglement to mediate the quantum computation. And also in 1992, Alexei Samsonovich, Alwyn Scott, who's a Mathematical Physicist, and myself published--actually Alexei did this as his thesis--simulated a microtubule with Frohlich resonance. Now, what Reimers et al did was they simulated a linear chain and somehow from a linear chain decided that you can't have--you can't have significant Frohlich effects in a three-dimensional microtubule. Maybe doing a three-dimension was too difficult for them. But what Alexei did was he took a two-dimensional sheet of tubulin and wrapped--treated as a torus, made boundary conditions so that this would match up with this, and this would match up with this, so you get a torus. I was able to run these simulations and he--and he found maxima and minima for Frohlich resonance energy at these--at these spots here, the black ones, which matched experimentally observed attachment patterns for microtubule-associated proteins. So, microtubule-associated proteins bind the microtubules at specific tubulin locations and their location--locations then determine the function of the microtubules in terms of the architecture and regulating the synapses and doing what microtubules need to do. So this is a way of representing and processing information. So in 1992, Alexei had shown this and they unaware of this paper which I think refutes their contention also. It was much more elegant than what they did, looking at a single tubulin chain. So this is now what we're thinking that a tubulin qubit might look like oscillating at 8 megahertz, as Jirí said or maybe it was [INDISTINCT], that Frohlich originally said gigahertz, but microtubules are big so it makes sense to me perfectly in some sense that they're slower and 8 megahertz is fine and the number of operations per second has reduced, but that's fine. We got plenty of that to play with. So--but the 8 megahertz is interesting for another reason. So I'm going to digress a minute. And here's [INDISTINCT] picture which I like very much. And so the 8 megahertz would be--and another megahertz would be in the--in the microtubule wall itself; not in the water, not in the C-terminal, but the microtubule itself and, interestingly, that's in the--that's in the RF for electromagnetic, but it's also in the ultrasound if you took--look--just think of mechanical vibration, mechanical oscillations. And recently there's been a movement looking at--effects on the brain on cognition and consciousness of transcranial therapies, transcranial magnetic stimulation for depression, for all kinds of things, transcranial electrical stimulation to enhance memory, and transcranial ultrasound using vibrational energy into the head through the skull showing effects in the brain. Now this was shown--behavioral effects was shown, but nobody was sure whether the--whether it was actually getting into the brain and having none thermal effects. And this guy William J. Tyler at ASU, Arizona State University, did a very good study with rats or mice, I guess, where they put electrodes in the mice, let them recover, they're walking around, they can record from the electrodes. They then did transcranial ultrasound. And in addition to behavioral effects saw interesting electrophysiological effects so proving that the ultrasound is having physiological effects in the brain. And they've got a grant from the government for some kind of mind control in--I'm not sure what they're trying to do actually, but it's--I think they have good intentions. But in their--I hope. And also they've spun off a company. And if you look at their prospectus, they're promising beneficial effects of transcranial ultrasound on about any psychiatric diagnosis you can think of, including from depression to memory problems to what have you. So anyway, we at the University of Arizona, including my colleague Chris Duffield who's here, have started our own studies. So I'm an Anesthesiologist, we skipped the rat business and we're going right to humans. And we've tried it on ourselves actually. It's a very interesting effect on your consciousness. Let's just put it that way. And so we think--we think people will like it. And we've got a proposal in for--we've got a proposal in for effects on chronic pain. And after that, we're going to put it--that's almost through, it looks pretty good for chronic pain patients. And after that we're going to try memory, effects on memory because as [INDISTINCT] showed, stimulating at 8 megahertz causes microtubules to polymerize and grow so you could enhance synaptic plasticity, you could enhance turn over, and enhance learning, and it could be a treatment for Alzheimer's disease. So I think this is a very, very promising new technology. And we're going to have a session on this as Hartmut mentioned, I co-organize a conference every year--every other year, it's in Tucson, and in the odd number years it's elsewhere in the world. So next May it'll be in Stockholm, Sweden, "Toward a Science of Consciousness: Brain, Mind, and Reality." And one of the sessions will be on transcranial therapies. We have Allan Snyder talking about--well, he's done a lot of transcranial magnetic, but he's going to talk about transcranial electrical. Tyler's going to talk about transcranial ultrasound and Wasserman from NIH will talk about transcranial magnetic stimulation. We're also going to have a session on Quantum Biology and several other interesting sessions. And I've left some flyers over there for you to take if you're interested. So conclusions; the objections to Orch OR by McKemmish et al and Reimers et al are invalid as far as I can tell. They didn't lay a glove on us as far as I can tell. Maybe I missed something, but they whiffed. There's nothing in there that's threatening in any way. Of course I'm biased, but you know. So Orch OR remains viable and finally it is far better to be criticized than ignored. Thank you very much. >> Okay, we have time for some questions. Could you use the microphone, please? >> Only a comment. Only a comment. The mistake of Dr. Reimers and Dr. McKemmish was the following two mistakes: First was, that they neglected to [INDISTINCT] linearity. And if you neglect [INDISTINCT] linearity, you must strongly [INDISTINCT] linearity. System can only be described linearly. You must have no mistake. And the second mistake, they neglected the fact that oscillators maybe normal, this--than [INDISTINCT] more than [INDISTINCT]. Oscillators critically damped, and oscillators over damped and then if you have over damped oscillators then the coherent time of which I spoke here about one microsecond, decreases to ten times [INDISTINCT] for instance or something like this. This is two mistakes in their work. The second point, I would like to say something that--concerning schizophrenia which is somehow is connected to this--to this--your work. A physician in our country measured immunoreactions by some special antigen and found that in illnesses that have disturbed, somehow, mitochondria function, there is a special reaction. And the reaction is the same in cancer for instance, in heart failures, in schizophrenia, and in predetermined [INDISTINCT] on the--very, very and so on predetermined. And in all these cases very likely, also in schizophrenia, is disturbed function of mitochondria therefore somehow that system that supply energy to oscillating system in the--in the cells which belong to biological activity. Thank you. >> Okay, thank you. Any more questions? >> HAMEROFF: Let me just say that Jirí's work is a huge help to us, number one, showing the eight megahertz coherence and also showing where the energy comes for the--for the excitations from the mitochondria without requiring GTP because that's--that was a critical issue as what's--I mean, Frohlich's [INDISTINCT] the heat bath, but that's a little bit vague. But showing that it's coming from mitochondria is very, very helpful and I think a great discovery. >> Thanks. It's always nice to hear it tour de force and that was really well done. And I had a question. If you think--I like to think of a cell as a four bit microprocessor or eight bit because it's hard to do much more unless you quantum it. And if you're a cell and somehow something signals you that you have to now be differentiated to a certain task, there's a complex set of subroutines. You have to know somehow how to extract if you're going to deliver work for example. So for a long time, I wondered, well if I'm going to discover, I have to do a subroutine, I need a bit to store to know that I'm now--I need something that works with the rest of the complex informatics which what you've shown with the CaMKII sounds like the perfect kind of a cellular memory. >> HAMEROFF: Yeah. >> I remember in 1999, you and Jack and Porter published a paper that also suggested, with some fascinating diagrams, that the microtubules of a cell can form electrical circuits also. And I just--I'm thinking, in addition to all the quantum discovery, if you're not showing us how there is a--almost a perfect microprocessor memory state in cells. >> HAMEROFF: Yeah. Well thank you. Yeah. As I said, I worked 20--for 20 years on the idea of microtubules as classical computers. I don't even think about quantum. And I only went to the quantum business when somebody stymied me with a question, "Okay, wise guy, how's that going to explain," you know, "how we have feelings and feel joy and redness and, you know, the heart—what's now known as the heart problem in consciousness research?" And at that time I read Roger's book, "The Emperors New Mind," which was beautiful for many reasons and he had a mechanism, but he didn't have a structure. I had a structure, but didn't have a mechanism. So I--you know, I wrote to him and he immediately got it. He immediately saw that microtubules could be, you know, what he needed and, you know, we teamed up. And we're kind of an odd couple. We're very different in almost every way, but he's a wonderful guy and, you know, he's still very interested. He's gone off to worry about, you know, the origin of the universe and stuff like that, but he's still very keen on this stuff and we stay in touch and he'll be doubling back into this. But microtubule is classical information processing, mediating differentiation and memory. And I don't know--maybe Jack knows whether CaMKII is found widely in other cells other than neurons--I think it is--and can mediate all kinds of things in terms of memory. There's got to be a memory site within. And it's not just a bit, each tubulin can be more than just two states. It can--it can be--it can be post-translationally modified. It can be tyrosinated. There's--the potential for information processing in microtubules is vast even in one cell. >> Any idea how fast? >> HAMEROFF: How fast what? >> A memory can be stored and retrieved? >> HAMEROFF: Well, within seconds the CaMKII is distributed widely throughout the dendrites of one neuron and then--and neighboring neurons. And if you stimulate simultaneously, it can go widely throughout the brain, so within seconds. Now, whether that's--you know, I think of it as probably a short term memory, not immediate consciousness because I think the--that's a little bit different. But the storage is seconds to minutes at the most. So--and then the retrieval can be--can be pretty fast. >> So one more question. Just guessing, is it possible the short-term memory can be stored one way and then we go to sleep and it goes to long term and... >> HAMEROFF: Yeah. Well there's a lot about that in terms of--when you go to sleep, the consolidated memory. And it involves, you know, hippocampal gamma oscillations and stuff like that. So that may involve some kind of hardwiring. Maybe it's transferred from the microtubules to the neurofilaments which are way more stable than even microtubules or maybe there is--there is a posttranslational modification, you know, enzymatic or non-enzymatic changes in the tubulin so it becomes more hardwired and that could be what memory consolidation is. >> Well, for what it's worth, keep it up. >> HAMEROFF: Thank you. >> Okay. Well, thank you. I'd like to thank Stuart and... >> HAMEROFF: Thank you.