C Vabvabc o0ci1c425

similarly,

This effectively concludes the teleportation of the state of MT A to MT C with one caveat.

Step 3: There is a probabilistic nature to this process, which means that MT C may receive the exact copy of the state of MT A, i.e. + (C)) or it may receive a state that is a unitary transformation away from the original (A)) (one of the other three possibilities: -) or \&±). MT C can reproduce the state of MT A if there is a "hardwired" condition so that when MT C receives + (C)) it does nothing further, yet if it receives one of the other three, it performs the correct unitary transformation to obtain the correct state from A. This "hardwired" behavior can be implemented through the use of codes, not unlike the Koruga bioinformation [62] code that MTs seem to follow. In principle, the exact state correspondence may not even have any significance and instead the information could be encoded in the frequency of transferred states, similar to information being encoded in the frequency of action potentials and not their shapes.

Teleportation is an entirely nonclassical phenomenon and has been experimentally demonstrated in matter and light states and combinations of these (see Sect. 4.5). Currently, the scientific consensus is that teleportation is impossible without entanglement although this may change in the future. Biological teleportation as described above, can be imagined as the basis of intra- and intercellular correlation that leads to yoked function (e. g. intra-cellularly during translation and intercellularly during yoked neuron firing). Experiments to check for such teleportation of states can be designed based on the surface plasmon resonance (SPR) principle [105] as applied to sheets of polymerized tubulin immobilized on a metal film (Sects. 4.4 and 4.5).

Fig. 4.2. Schematic of quantum teleportation of dipole states. (Taken from [81]) MT (a) sends its state (represented by the yellow cross) to MT (c) without any transfer or mass or energy. Both MT (a) and MT (c) are entangled with MT (b). Entanglement represented by the presence of connecting MAPs (green)

Fig. 4.2. Schematic of quantum teleportation of dipole states. (Taken from [81]) MT (a) sends its state (represented by the yellow cross) to MT (c) without any transfer or mass or energy. Both MT (a) and MT (c) are entangled with MT (b). Entanglement represented by the presence of connecting MAPs (green)

A graphical representation of biological quantum teleportation of dipole moment states is presented in Fig. 4.2.

Information Processing by Biopolymers and the Guitar String Model (GSM)

In the quantum-mechanical scenario for MT dynamics discussed above, as suggested in [80], a quantum-hologram picture for information processing of MT networks emerges. Further, the existence of solitonic quantum-coherent states along the MT dimer walls implies a role for these biological entities as logic gates [82]. Consider, for instance, a node (junction) of three MTs connected by microtubule-associated proteins (MAPs) see Fig. 4.3. The quantum nature of the coherent states makes the junction interaction probabilistic. Therefore, at tube junctions one is facing a probabilistic Boolean interaction. The probability of having a solitonic coherent state in a MT branch does depend on its geometric characteristics (such as length). By modulating the length of the tubes and the binding sites of the MAPs a bias can be introduced between bit states that can affect the probabilistic final outcomes. This has obvious implications for information processing by MT networks.

4 Quantum Effects in Cytoskeletal Proteins 119 (b)

Fig. 4.3. An XOR (exclusive-or) logic gate. (Taken from [81]) "0" ("1") is represented by absence (presence) of solitonic kink wave of dipole moment flips indicated in yellow. (a) Input MT, (b) Output MT, (c) A MAP transmitting a solitonic kink wave, (d) a "quiet" MAP. MT (a) has two solitons traveling, encountering two MAPs that transmit both solitons to MT (b). In this hypothetical scenario, the solitons arrive out of phase at MT (b) and cancel each other out. The truth table for XOR reads: 0,0^0; 0,1^1; 1,0^1, 1,1^0 and can be realized by an MT arrangement if the MAPs are arranged such that each can transmit a soliton independently but if they both transmit the solitons cancel out, i. e. the two MAPs must be an odd number of dimers apart on MT(a)

Fig. 4.3. An XOR (exclusive-or) logic gate. (Taken from [81]) "0" ("1") is represented by absence (presence) of solitonic kink wave of dipole moment flips indicated in yellow. (a) Input MT, (b) Output MT, (c) A MAP transmitting a solitonic kink wave, (d) a "quiet" MAP. MT (a) has two solitons traveling, encountering two MAPs that transmit both solitons to MT (b). In this hypothetical scenario, the solitons arrive out of phase at MT (b) and cancel each other out. The truth table for XOR reads: 0,0^0; 0,1^1; 1,0^1, 1,1^0 and can be realized by an MT arrangement if the MAPs are arranged such that each can transmit a soliton independently but if they both transmit the solitons cancel out, i. e. the two MAPs must be an odd number of dimers apart on MT(a)

Such a binary information system can then provide the basic substrate for quantum information processing inside a (not exclusively neural) cell. In a typical MT network, there may be about 1012 tubulin dimers. Although such a number is large, as discussed earlier there may be subtle "shielding" mechanisms at play. The above scenario is not necessarily quantum in nature. An essentially identical argument can be made for information processing via waves of dipole flips or just momentum transfer as a result of propagating conformational changes.

This suggests an obvious model for encoding information in a network of MTs that we shall call the "guitar string model" to emphasize the analogy to the way six guitar strings (the MTs) can by clamped by four fingers (MAPs) at different nodes to generate hundreds of different chords (engrams). If propagating dipole moment flips are indeed carrying signals inside the cell then the nodes in the network can affect this propagation in a large variety of ways. A limited set of MTs with a limited set of MAP binding sites can have a very large set of engrams. This also suggests a way for new memories to form and old ones to be erased by simply changing the distribution of MAPs.

4.2.4 Conclusions

If it is experimentally confirmed that treating MTs as QED cavities is a fair approximation to their behavior, one can propose that nature has provided us with the necessary structures (microtubules) to operate as the basic substrate for quantum computation either in vivo, e. g. intracellularly, or in vitro, e. g. in fabricated bioqubit circuits. Such a development would pave the way to construct quantum computers by using quantum computers by using micro-tubules as building blocks, in much the same way as QED cavities in quantum optics are currently being used in successful attempts at implementing qubits and gates [126]. Detecting quantum behavior at this level would undoubtedly advance attempts at implicating quantum physics.

4.3 Tau Accumulation in Drosophila Mushroom Body Neurons Results in Memory Impairment

4.3.1 Introduction

In this section we summarize and attempt a "physicist-friendly" account of the neurobiological results obtained by our group and published in [86].

In order to test some of the predictions of the models discussed in Sects. 4.1 and 4.2, an in-vivo neurobiological behavioral study was undertaken. The goal was to experimentally investigate whether memory is affected by perturbations in the microtubular (MT) cytoskeleton. Associative olfactory learning and memory were the types of memory accessible to us with the transgenic Drosophila fruitfly behavior analysis system. We tried disturbing the fly MTs as little as possible, avoiding perturbing the cytoskeleton by formation of such large protein aggregates as neurofibrillary tangles (NFTs) that could effectively "strangle" the neuron, disrupting or even stopping intracel-lular (axonal) transport. In addition, NFTs and/or amyloid or senile plaques (APs or SPs) have been unequivocally shown to contribute to neurodegeneration and eventual neuronal death and it is reasonable to expect a dying neuron to dysfunction, regardless of the state of its MTs. We also avoided causing any developmental problems by selecting gene promoters (drivers) with appropriate temporal activity.

Since this is a contribution addressed mostly to physicists an effort has been made to explain potentially unfamiliar biological terms and procedures. Following established standards in genetics, small case italics such as tau indicate the gene that codes for the protein TAU indicated in capitals. Strains or lines of transgenic animals (animals that have been genetically manipulated and contain extra genes) are named somewhat arbitrarily so here we use b-, h-, or d-, indicating bovine- human- or Drosophila- (native) derived-genes, while a few letters identify the source.

Microtubule-associated protein (MAP) TAU has long being implicated in the encoding of human memory and it has been shown that mutations in the human NC-17 tau gene are one of the causes of Alzheimer's Disease [95, 99, 138]. For this reason, NFT and SP/AP formation have been the main focus of studies of tauopathies in animal models. For instance, transgenic mice with frontotemporal dementias with Parkinsonism (FTDP 17) mutations develop NFTs and neurodegeneration accompanied by motor deficits [47, 71]. Expression of human wild type and FTDP-17-linked mutations in Drosophila results in age-dependent neurodegeneration without NFTs [141] except when wild-type TAU was phosphorylated by overexpressed Drosophila glycogen synthase kinase-3 [50]. Mice carrying mutated tau, presenilin 1 and aplha-beta peptide precursor (APP) transgenes show synaptic dysfunction before the development of NFTs or amyloid plaques. From these and other studies it seems that tauopathy-caused deficits in memory appear even without NFTs or SPs/APs although frequently, at least NFTs do eventually appear in the late stages of the disease.

For NFTs to form there must be a situation of elevated TAU accumulation (in a nonfilamentous or "pretangle" state) in the affected neurons. Such a condition has been suggested as the underlying cause of pre-neurodegeneration cognitive symptoms such as memory loss [15, 47] and our research experimentally addresses the question of the effect of elevated pretangle state TAU in Drosophila mushroom bodies and we propose a connection between the observed effects and theoretical models of cytoskeletal function.

4.3.2 Drosophila

The Drosophila Melanogaster fruit fly has long been a favorite of experimental behavioral neurobiologists for numerous reasons including its relatively simple genetic makeup and quick generation time, powerful classical and molecular genetics and the animal's ability to learn and remember a variety of tasks.

To illustrate our approach and choice of Drosophila more fully, our initial experimental design will be briefly described here. Drosophila was selected as the ideal system for investigating cytoskeletal involvement in learning and memory because we were to attempt to track an intraneuronal redistribution of MAP-2 and/or MAP TAU as a result of conditioning. This is a prediction of the GSM described in Sect. 4.2. In order to track a redistribution of MAPs inside neurons one must be able to differentiate between the various parts of the neuron such as the dendrites, axons, axonal projections and somata. In humans and other mammals, the neuronal organization is such that multiple neurons and neuronal types are involved in a given process forming an extensive complex network of axons and dendrites. As a result, it is particularly difficult to locate individual neurons' specific parts and stain selectively to track changes in distribution of a particular protein. In Drosophila, on the other hand, the neuronal organization is such that differentiation of subneu-ral parts is facilitated. For instance, neurons belonging to mushroom bodies (MBs are prominent structures in the Drosophila brain essential for olfactory learning and memory) represent a highly ordered, tightly and sequentially packed neuronal system where axonal projections (i.e. synaptic fields), dendrites and somata are macroscopically (on the order of |m) separated in ordered fiber bundles, see Fig. 4.4.

This provides a strong advantage for analysis of the results of expression of microtubule-associated proteins in specific neurons (e. g. those associated with a specific type of memory) but also within different parts of such neurons. For instance, a bulk redistribution of a certain MAP from the axons to the dendrites of the MB, presumably as a result of memory formation can, in principle, be tracked. This is in fact a prediction of the GSM for memory encoding since if the MAPs play the role of fingers on the guitar fret board and the various chords correspond to encoded information, acquisition of new information and memory would result in a redistribution of MAPs. Unfortunately, our preliminary experiments utilizing directed expression of chicken MAP-2 in MBs showed that either the resolution offered by existent anti-MAP antibodies was insufficient to decipher appreciable MAP redistribution and/or no such redistribution took place as a result of learning. The latter would be inconsistent with results obtained in rodents [143] that suggest a redistribution of MAP-2 resulting in accumulation in dendrites as a result of learning an auditory associative task.

Fig. 4.4. Fly mushroom bodies (MBs). Fly mushroom bodies are shown in paraffin frontal sections 5 |J,m thick, stained for LEO, a MB-specific protein

We therefore shifted our approach to determining the impact of MAP TAU overexpression on the ability of the animals to learn and retain memories. Although this is not as direct a test of the GSM it does provide a solid link between the microtubular cytoskeleton and memory retrieval and stability as will be argued in this section.

4.3.3 Genetic Engineering

We induced the expression of vertebrate (human and bovine) tau genes, producing TAU protein in specific tissues and at specific times in Drosophila using the method of directed gene expression.

Directed Expression

Directed gene expression rests on the principle of obtaining two genetically manipulated (transgenic) lines, the first of which contains the gene to be expressed, fused to and under the direction of an upstream activating sequence (UAS). This UAS promoter is activated by the presence of its unique, selective and specific activator protein GAL4. To generate transgenic lines expressing GAL4 in a cell or tissue specific pattern, the GAL4 gene is inserted randomly into the fly's genome, thus driven in its expression from various genomic enhancers. A GAL4 target gene (UAS-tau) will remain silent in the absence of GAL4. To activate the target gene, the flies carrying the UAS-tau are crossed to flies expressing GAL4 at specific tissues and at specific times in the animal's development (see Fig. 4.5). To eliminate potential complications arising from expression of TAU in the embryonic and developing nervous system, we selected strains expressing GAL4 in late pupal and adult mushroom body neurons [3] only by utilizing the MB drivers c492 and c772.

Fig. 4.5. Upstream activating sequence and target gene

Transcriptional aaivalionofTAU-gene

Fig. 4.5. Upstream activating sequence and target gene

Coimmunoprecipitation

Note that past studies have shown that mere accumulation of non-Drosophila proteins such as ,3-galactosidase [124, 125], and GAL4, or Drosophila proteins [30, 83] that are not MT specific in MB neurons, do not cause any behavioral deficits. Therefore, once TAU is shown to bind to MB MTs, any effects of TAU accumulation in MBs can be taken as specific to TAU. Also note that the mere presence of two proteins in the same tissue does not necessarily mean they are bound together so although we expected h- and b- TAU to bind to MTs we performed a coimmunoprecipitation study. The Drosophila protein (dTAU), contains four putative tubulin-binding repeats [43]. They exhibit 42% and 46% identity (62% and 66% similarity)2 with the respective sequence of bTAU and hTAU [43-45]. To determine whether this sequence conservation among vertebrate and Drosophila TAU also signals a functional conservation (i. e. if all types retain their MT binding sites intact), antitubulin antibodies were used in immunoprecipitation experiments from head lysates of btau-expressing animals and controls. We found that bTAU coimmuno-precipitates with Drosophila tubulin, indicating that the vertebrate protein is capable of binding Drosophila microtubules. Figures and extensive details are presented in [86].

There is one obvious problem in assuming that we have just substituted "more of the same" in the fly's mushroom bodies because dTAU lacks the amino-terminal extension of vertebrate TAU and this suggests that the conformation of vertebrate TAU will be somewhat different from the Drosophila protein despite its microtubule-binding ability. This, however, does not affect the conclusions of this study, as will be illustrated later.

Collectively, the ability to bind Drosophila tubulin in head lysates and its preferential accumulation within the mushroom bodies indicate that bTAU, and by virtue of its high degree of identity hTAU, bind to the microtubular cytoskeleton within these neurons. Therefore, in the mushroom body neurons of the three tau transgenics investigated, the microtubular cytoskeleton is likely burdened with excess TAU.

All strains were normalized to an isogenic (i. e. genetically identical) w1118 strain.3 To obtain flies for behavioral analyses, c772 and c492 homozygotes were crossed to UAS-btau, UAS-htauwtl, UAS-dtau 4 and UAS-dtaul homozygotes (see separate section on dtau below) and the progeny was collected and tested 3-5 days after emergence. Similarly, the UAS-btau, UAS-htauwtl,

2 Identity is defined as absolute conservation of the amino acid sequence between two proteins, while similarity is conservation of type (e. g. exchanging one acidic amino acid for another acidic preserves similarity)

3 Isogenic lines are strains of identical genetic background. w1118 was chosen to represent the wild-type genotype. The transgene of interest was bound to redeye phenotype and the trangenic flies were crossed to w1118 for seven generations (keeping only the red eyes) thus normalizing the genetic background and avoiding contamination.

UAS- dtau 4 and UAS- dtaul homozygotes were crossed tow1118, the line not containing any drivers (and thus one does not expect to see any extra tau expression) to obtain heterozygotes used as controls.

In addition to the pattern of expression of a gene, it is important to also quantify the amount of protein that is being created. To investigate the relative level of TAU accumulation within adult fly heads we performed semiquantitative Western 6 lot analyses (see [86] for details). We determined that bovine TAU was present in head lysates of animals that had the tau transgenes and the MB drivers, but not in parental strains as was expected. The level of bTAU protein did not appear to change significantly over a three-

Fig. 4.6. (Taken from [86]) To determine whether the vertebrate (bovine) TAU binds onto Drosophila microtubules, extracts were prepared from the brains of flies that express UAS-btau under the c772 (lanes 2 and 3) and the c492 (lane 5 drivers) under mild conditions that would not disrupt TAU/tubulin complexes. A monoclonal antibody against tubulin was added to the mix and then precipitated using standard techniques (coimmunoprecipitation). The resulting complexes were resolved in a denaturing polyacrylamide gel blotted onto a nylon membrane and the presence of bovine TAU was investigated using a monoclonal antibody. Lane 1 does not contain bTAU because the extracts were prepared from heads of animals that contained only the btau transgene without the driver and therefore do not accumulate the vertebrate protein. Lane 2 represents extracts from bTAU accumulating animals under the c772 driver that were not subjected to coimmunoprecipitation. Therefore the characteristic bTAU band serves as a marker of the size and the amount of total bTAU in the extract. Lane 3 represents that products of coimmunoprecipitation from an equal amount of protein as in lane 2. Detection of a bTAU-specific band demonstrates that most of the vertebrate protein (compare the intensities the band in lane 2 indicating the total amount of bTAU in the head extract with the intensity of the band in lane 3 indicating the amount of bTAU that is complexed with Drosophila tubulin in the extract) exists in a complex with the Drosophila tubulin. Similar results were obtained with the c492 driver (lane 5). Notably, these bTAU specific bands were absent in the negative controls (extracts from animals not expressing bTAU because they do not carry the UAS-btau transgene, but only the drivers (lanes 4 and 6)

week period, indicated by densitometric quantification of results from three independent experiments. Similarly, the level of hTAU appeared relatively constant (data not shown).

Although it would be desirable to know exactly how much more TAU than normal is present, the techniques we used for quantifying the presence of b and h protein relied on mono- or polyclonal antibody binding and densitometry and thus are inherently difficult to normalize.

In principle, it would be possible to refine these findings with such elaborate methods as ion trapping and matrix-assisted laser-directed ionization and time-of-flight spectroscopy. Even though we did not attempt such a high degree of quantification, we are confident that the transgenic animals did express a significantly higher level of TAU in their mushroom bodies and this is sufficient for the scope of our study.

4.3.4 Conditioning

Drosophila fruit flies are naturally attracted or repulsed with a variety of affinities by different odors. We followed two standard negatively reinforced associative learning paradigms that essentially generalize the Pavlovian conditioning protocol by coupling aversive odors as conditioned stimuli (CS+ and CS-), and electric shock as the unconditioned stimulus (US). In this way, olfactory cues are coupled with electric shock to condition the flies to avoid the odorant associated with the negative reinforcer. These conditioning protocols for Drosophila were initially developed by Tully and Quinn [134] and modified by Skoulakis et al. [106, 125]. We used two aversive odorants: 3-octanol (OCT) and Benzaldehyde (BNZ). The conditioning apparatus consists of a training chamber and a selection maze (see Fig. 4.7). The maze is normalized by adjusting the concentration of odorants. Once normalized, both wild-type (control) and transgenic naïve (i. e. untrained) flies choose to enter one of two identical tubes smelling of OCT and BNZ, respectively, with a probability of 50% (as they avoid both odors equally). Because the earliest possible time that we can test the animals past the CS+ and US presentation is 180-200 s, our measurements cannot differentiate between "acquisition" and "3-minute memory". This earliest performance assessment will be referred to as "learning".

Conditioning of the flies in the LONG training protocol takes place as follows. A batch of wild-type, naïve flies (numbering between 50 and 60) are collected under light anesthesia (using CO2) and 12-24 h later are left in the dark for one to two hours. The entire conditioning procedure takes place in a temperature- and humidity-controlled darkroom under red illumination in order to isolate the effects of olfactory stimulation from visual stimulation (flies have been shown to react least to red light). Once the flies have been acclimated to the darkroom, half are inserted into conditioning chamber A. The cylindrical wall of the chamber is covered by a grid of two interspersed conducting electrodes spaced such that at least two of the fly's six legs must be

TESTING:

VACUUM t

Was this article helpful?

0 0

Post a comment