Fig. 4.7. Conditioning apparatus and training schedule. (Taken from [88])

in contact with two opposite voltage electrodes. The diameter of the chamber is chosen such as to prevent flies from hovering midair. In this way, the fly necessarily completes the circuit making it the path of least resistance. The electrodes are electrified by a signal generator set to 92.0 V. The flies receive eleven electric shocks once every five seconds. During this time, the chamber is filled with air containing OCT. The flies are given 30 s to rest while the air is being cleared of odorants and are then given the opposite (control) odorant (in this case BNZ) for another minute in the absence of electrical shocks. A rest period of 30 s follows after which the flies are tested for acquisition of memory by being inserted into the selection maze and given the choice of entering a chamber smelling of OCT or an identical one smelling of BNZ. For control and consistency purposes, the experiment is done simultaneously in an identical apparatus B with the shock-associated and control odors reversed. We define a conditioned or "trained" fly as one that has chosen to go into the chamber filled with the control odor after given the choice for 90 s.

It is observed that following training, a large percentage of wild-type flies choose to avoid the smell that was present when they received the electric shocks. The percentage is calculated as a normalized performance index (PI)

where PI = (tramedo™tramed) x 100%. Typical PI values for wild-type flies were between 75 and 90%, giving us confidence that the flies have learned to associate the stimuli. The general procedure described above is a typical associative learning Pavlovian conditioning paradigm for behavioral experiments appropriate for a variety of animals and more details can be found in the literature [106, 125]. One improvement that was discovered (and increased wild-type PI to about 90%) was that it was possible to collect the flies without subjecting them to CO2 anesthesia after conditioning by simply tapping the chambers sharply so that the flies entered a test tube through a funnel by their own inertia. The mechanical shock associated with such tapping was shown to have no effect on the flies' PI, while the (light) anesthesia immediately following conditioning as well as the presence of a naturally leaky CO2 tank in the darkroom was known to compromise PI scores.

4.3.5 Controls

During physics experimentation, background measurements play a significant role in determining the signal. Similarly, a dominant theme in biological behavioral research of the type described here is that of a set of measurements collectively called controls. When results such as, for instance, a decrement in learning and memory exhibited by transgenic animals are presented, they must always be quantified with respect to the equivalent in the control (nontransgenic or wild-type animal).

For our behavioral analyses to have any significance we had to first determine that the flies expressing foreign TAU or overexpressing native TAU were not affected in ways unrelated to learning and memory. Thus we had to assess the experimental flies' task-relevant sensory behaviors and olfactory acuity to an attractive odor in addition to testing their ability to feel and avoid electric shock and to detect and avoid aversive odorants. Experience-dependent nonassociative behavior was also tested by examining the effects of pre-exposure to odorant plus shock since pre-exposure to electric shock and one odorant tends to decrease the animals' ability to perform well in associative learning tasks that depend on electric shock and another odorant.

We also had to determine that our flies were viable and no neurodegeneration took place as the result of expression of the transgenes. Note that in all these control experiments, the main principle is to test the transgenic against the control, for instance, an absolute decrement in PI is not indicative of a TAU-mediated effect if it is mirrored identically in control animals.

Finally, one must be certain that the observed effects are TAU specific, i. e. that other proteins that do not bind the microtubular cytoskeleton and are (over)expressed with the same drivers, result in no memory effects. This was done and was discussed in Sect. 4.3.4.


Sensory control experiments to ascertain that the transgenic flies retained their mechanosensory abilities (to feel and avoid pain caused by electric shock) were performed as described in [106, 125]. Avoidance of electrified grids kept at 92 V (normal US stimulus), or 45 V was not different between tau -expressing animals and controls, indicating lack of mechanosensory deficits due to TAU accumulation. The two levels of voltage were used as a further refining mechanism to help find any mechanosensory effects that our transgene expression may have had (e. g. it may have made the flies less sensitive but still able to sense 92 V).

Olfactory Acuity Pre-Exposure

Although, as expected CS+/US pre-exposure significantly reduced subsequent avoidance of the complementary odor, tau-expressing animals and controls exhibited equal decrements (Fig. 4.8A). Therefore, TAU accumulation did not cause differential responses to odor-shock pre-exposure and we can conclude that TAU accumulation does not affect experience-dependent, nonassociative tasks.

Attractive and Aversive Odors

Both control and transgenic tau-expressing animals avoided equally the aver-sive odors benzaldehyde and 3-octanol (CS) at two different odor concentrations given the choice of fresh air. These results indicate that TAU accumulation in mushroom body and other central brain neurons described above, did not result in deficits in sensory abilities necessary for olfactory conditioning. In addition, we tested the response of btau and htau-expressing animals relative to controls to the attractive odor geraniol (GER) in olfactory trap assays [125]. Though this odor is not task relevant, it provided an independent measure of olfactory acuity towards a qualitatively different odor.

As shown in Table 1 of [86] the performance of tau-expressing animals was not significantly different from controls for these tasks. Collectively, the results of these olfactory control experiments support the conclusion that despite the accumulation of TAU in antennal lobe neurons (that are used as olfactory sensors by the fly) btau- and htau-expressing animals retained their normal olfactory responses to the odors tested [86].

Viability of Transgenics

To assess the effects of extra TAU in MBs on the learning and memory capabilities of flies convincingly, we had to establish that our flies would be reasonably healthy during their conditioning and testing. We hypothesized that accumulation of TAU within the mushroom bodies would not affect survival because these neurons are dispensable for viability [24]. However, the

effect on fly longevity of TAU accumulation in additional neurons where c772 and c492 drivers were determined to be active was unknown. In addition, it has been recently shown that expression of human wild-type and mutant TAU proteins in the entire Drosophila nervous system (pan-neural expression), or targeted expression in cholinergic neurons, results in neurodegeneration and

Fig. 4.8. Results (taken from [86]) (A) Nonassociative pre-exposure effect. 1. Avoidance of benzaldehyde after pre-exposure to full strength octanol and 90-V electric shock (filled bars) in comparison to avoidance without such preexposure (open bars) (n > 7). ANOVA revealed significant effects of treatment (F(i,78) = 13.784, p(0.005), but not for genotype, both in pre-exposed and non-pre-exposed animals. 2. Avoidance of octanol after pre-exposure to benzaldehyde and 90-V electric shock (filled bars) in comparison to octanol avoidance without pre-exposure (open bars) (n > 7). ANOVA revealed significant effects of treatment (F(i,84) = 14.026,p{0.005), but not for genotype, both in pre-exposed and non-pre-exposed animals. (B) Olfactory memory after LONG paradigm conditioning. The mean performance index ± SEM of c492/+, c772/+, UAS-dtau4 /+ (open bars) and c492/+; UAS-dtau4 /+ and c772/+; UAS-dtau4 /+ (filled bars) are shown (n > 9). Two-way ANOVA revealed significant effects of genotype [(F(4,52) = 14.687, p(0.005) immediate (3-minute), and (F(4,49) = 9.327, p(0.005) 1.5h]. Subsequent Dunnett's tests for each time interval did not reveal significant differences in performance among the c492/+, c772/+, UAS-dtau4/+ control strains or between the c492/+; UAS-dtau4 /+ and c772; UAS-dtau4/+ heterozygotes. However, the differences between c492/+; UAS-dtau4 /+ and c772; UAS-dtau4/+ heterozygotes and the control strains were highly significant (p(0.001) for immediate memory and 1.5 h memories. (C) Performance of c492/+; UAS-dtaul /+ and c772; UAS-dtaul/+ heterozygotes with or without induction at 29°C after LONG conditioning. The average performance (PI± SEM) of animals raised at 23-24°C for control strains (c492/+, c772/+, UAS-dtaul/+) is indicated by open bars and c492/+; UAS-dtaul /+ and c772; UAS-dtaul /+ heterozygotes with gray filled bars. The performance of animals raised at 23-24° C and subsequently induced at 29° C for 48-52 h prior to behavioral experiments is indicated by the stippled bars for controls and the black-filled bars for c492/+; UAS-dtaul /+ and c772; UAS-dtaul /+ heterozygotes. Two-way ANOVA indicated significant effects of genotype [(F(4,44) = 8.287, p(0.005) for 23-24°C animals and (F(4,48) = 10.016, p(0.005) for animals induced at 29°C]. Subsequent Dunnett's tests revealed significant differences between the performance of c772; UAS-dtaul/+ heterozygotes and all control strains, as well as c492/+; UAS-dtaul/+ when uninduced (p(0.001). In contrast, both c772; UAS-dtaul/+ and c492/+; UAS-dtaul/+ heterozygotes were different from controls when the animals were induced at 29°C

premature death of adult flies [141]. In order to determine whether TAU expression in adult mushroom bodies and the other brain neurons described above affects the flies' viability, we evaluated the survival of btau-, htauwtl -and dtaw-expressing flies over a period of 21 days posteclosion.4

Because both male and female animals are used for our behavioral experiments, we used mixed-sex populations to evaluate survival, unlike previous studies [17, 141]. We concluded that expression of vertebrate or Drosophila tau in adult mushroom body and other brain neurons did not result in decreased survival [86]. No overt differences between control strains and transgenics were observed for a limited set of animals that were evaluated for

4 Eclosion refers to the adult fly emerging from the pupal case (cocoon).

viability for two additional weeks (data not shown). Furthermore, TAU accumulation did not appear to result in gross morphological differences, or decreased fecundity and vigor from control strains.

These results indicate that TAU accumulation within the MB and other neurons of the adult brain does not precipitate the neurodegeneration-dependent decrease in survival observed with pan-neural expression of human mutant TAU proteins throughout development [141].

Neuroanatomy and Histology

Although the mushroom bodies are not essential for viability, we expected that degeneration of these neurons would severely impair behavioral neuro-plasticity [24]. To determine whether TAU accumulation in adult mushroom bodies causes their degeneration, we histologically investigated the brain neu-roanatomy of animals that expressed the tau transgenes. Because in past studies the severity of neurodegeneration was observed to increase with age and accumulation of TAU [24], we focused on 21-day-old animals. Using a semiquantitative western blot [86] we concluded it is unlikely that any degeneration in older flies is the result of progressively increasing amounts of TAU in mushroom body neurons. Note here that although as previously mentioned, densitometry and immunohistochemistry are by nature hard to normalize, they do give reasonably accurate relative results. Thus we were able to determine that no more TAU was present in older flies even though we could not tell exactly how much overall TAU there was. Staining with various mushroom body antigenic markers [23] did not reveal detectable morphological anomalies in 21-day-old transgenic tau-expressing flies compared to similarly aged, or 2-3-day-old controls.

4.3.6 Results

The results of this study can be summarized as follows. Vertebrate (bovine and human) as well as native TAU accumulation in mushroom bodies (implying TAU bound to MTs) of adult flies results in associative olfactory learning and memory deficits.

Vertebrate TAU-Expressing Flies

We had determined that transgenic animals under the c772 MB driver express higher levels of TAU so we used c772/+; htauwtl /+ heterozygotes in the analysis presented below, and similar results were obtained for c4922/+; htauwtl/+ heterozygotes in a limited set of experiments (data not shown).

To determine whether TAU accumulation in the mushroom bodies affected associative processes, we trained btau-, htau- and dtau-expressing animals and controls in the long version of a negatively reinforced, olfactory associative learning task as described above. c492/+; UAS-btauI/+, c772/+; UAS-btauI/+ and c772/+; UAS-htauwt1 /+ heterozygotes exhibited a highly significant 25-30% impairment in learning compared to controls (Fig. 4.8B, immediate). Similar results were obtained for driver X dtau and they are described in the literature [86]. These results demonstrated that TAU accumulation in the mushroom bodies strongly compromised behavioral neuro-plasticity underlying associative olfactory learning and memory.

To more closely examine the learning and memory deficits of btau-, htau-and dtau-expressing animals, we utilized the SHorT variant of associative olfactory training [10] performed by E.M.C. Skoulakis. Because the LONG paradigm utilizes a 60-s CS+ presentation concurrent with 11, 92 V electric shocks, the flies' performance represents learning from multiple rounds of what is referred to as "massed" CS/US pairing. On the other hand, in the SHorT paradigm, a 10-s CS+ presentation is coupled to a single 1.25-s, 92-V shock, allowing assessment after a single CS/US pairing [10, 19]. Furthermore, performance in SHorT training improves upon multiple pairings with a 15-min intertrial interval [10, 19]. This allows for a very fine-tuned experimental manipulation to produce equivalent learning in control and experimental animals, a necessary condition to investigate memory stability and retrieval properties.

The results in Fig. 4.8C demonstrate that a single CS/US pairing in btau-and htau-expressing animals yielded losses in learning scores of the order of nearly 50% relative to controls. As with controls, the performance of tau-expressing animals improved upon multiple CS/US pairings, indicating that the basic neuroplasticity mechanisms were at least operating in the right direction, but three CS/US pairings were necessary for tau-expressing animals to perform at the level reached by controls after only two pairings (Fig. 4.8C). This suggests that TAU accumulation causes either an impairment in the learning resulting by each CS/US pairing, or a compromise of memory stability, retrieval, or a combination of the two.

To distinguish between these three possibilities, we trained c772; UAS-btau and c772; UAS- htauwt1 heterozygotes to the same performance level as controls (3 pairings for tau-expressing animals and 2 for controls) and measured memory of the association after 30 min. The tau-expressing animals exhibited a significant decrease in 30-min memory, despite performing equivalently to controls immediately after training. This indicates that memory retrieval and/or stability were compromised in TAU-expressing animals. This result implicates TAU within mushroom body neurons to mechanisms that are key to memory stability and/or retrieval.

Since our coimmunoprecipitation experiments showed that all TAU tested did bind to MTs we can conclude that the behavioral deficits observed are the effect of burdening MTs with excessive TAU. This is in accord with what one would expect if the MTs were the first (or at least near the "front lines") of intercellular information-manipulation elements.

Integrating Results from h-, b-, diaw-Expressing Animals

Combining the results summarized above and those in [86] we are led to the conclusion that the decrements in learning and memory observed in btau-and htau-expressing animals were not caused by accumulation of a vertebrate protein, but rather by increasing the level of TAU in these neurons. Low levels of dtau transcription did not affect the performance of c492/+; UAS-dtau 1/+ animals. However, elevation of dtau transcription precipitated learning deficits similar to those observed with vertebrate tau and dtau4 transgenics [86]. These results strongly indicate that the associative learning and memory deficits in vertebrate tau - and dtau-expressing animals are very likely the direct result of elevated TAU accumulation within mushroom body neurons and not because of the conformational differences between vertebrate and Drosophila proteins.

Finally, another unlikely scenario that fits the data is that since the native and overexpressed dtau genes were in different parts of the genome, some role was played by the directed expression on the conformation of the extra dTAU resulting in a perturbation that made dTAU behave like its b and h analog. One way to rule this out would be to entirely knock out the native TAU gene and replace it with htau or btau, thus testing whether these will take up the role meant for dtau and thus possibly refining the findings reported here and also determining whether the hypothesis behind the "fetal, 4R" TAU versus the "adult 3R" TAU in Alzheimer's disease is correct.

4.3.7 Conclusions

Collectively, the results of the behavioral analyses suggest that the level of TAU within mushroom body neurons is essential for both olfactory learning elicited by each CS/US and memory retrieval or stability. The areas of significant homology between the vertebrate and Drosophila TAU are confined to the tubulin binding sites and the vertebrate protein appears to bind mi-crotubules in a way similar to the way the fly protein does. Taken together, the results strongly suggest that excess TAU binding to the neuronal micro-tubular cytoskeleton causes mushroom body neuron dysfunction exhibited as learning and memory deficits. This also indicates that although excessive TAU may not result in (immediate or medium-term) neurodegeneration, it is sufficient to cause significant decrements in associative learning and memory that may underlie the cognitive deficits observed early in human tauopathies such as Alzheimer's.

4.3.8 Discussion

The pretangle state of elevated tau has been the topic of limited study in the past (e.g. loss of TAU in axons and elevation in the somatodendritic compartment of neurons prior to tangle formation shown in humans and animal models [5, 15] but the possible effect of this state on neuroplasticity and possibly consciousness (if one is willing to make the leap of faith and extend these findings from fly to human), had not been previously explored. Similarly, splicing mutations that increase the level of 4R (fetal) TAU are the hallmark of many human tauopathies [35, 70]. It has been argued that accumulation of unbound TAU and subsequent NFT formation in human tauopathies may be the result of conformational changes [93]. However, the conformational differences between dTAU and its vertebrate homologs did not appear important in affecting learning and memory deficits in our study, at least at the level of resolution we could obtain. In contrast, the overall level of TAU, i. e. the MAP:MT stoichiometry within mushroom body neurons appeared to be of primary importance.

TAU accumulation in mushroom body neurons caused robust associative learning and memory deficits. It is surprising that within the resolution limits of our techniques, the deficits appeared confined to associative learning and memory and not to other experience-dependent olfactory processes. These results suggest that normal cytoskeletal-mediated processes, likely disrupted by excess TAU are necessary for neuroplasticity underlying associative functions. One possible overarching explanation of our findings involves the disruption of axonal transport of vesicles and is explored fully in [86]. An alternative, of interest to us as explorers of the QCI involves the roles for neuronal microtubules and their dynamic interaction with the proper ratio of MAPs in learning and memory discussed in Sect. 4.2 and [142, 143], or as neuronal computational elements proposed in [81]. These data are consistent with specific predictions of these models, including the GSM model that predicts that perturbations in the ratio of microtubule-binding proteins will precipitate learning and memory dysfunction and also with the general approach behind the dipole-dipole logic suggestion where extra TAU would alter the local dielectric constant k by virtue of its increased density

In summary, these results strongly suggest that the stoichiometry of TAU and microtubules within neurons is essential for behavioral neuroplasticity. Increasing the level of TAU within neurons precipitates deficits possibly due to inhibition of microtubule-dependent intraneuronal traffic, microtubule stability or interactive capacity. The strong behavioral effects indicate that directed TAU-accumulation within neurons can be used as a tool to disrupt and study neuronal function in general.

4.4 Refractometry, Surface Plasmon Resonance and Dielectric Spectroscopy of Tubulin and Microtubules

4.4.1 Theory of Dielectrics

Dielectric Properties of Polar Molecules

The dipole moment p is defined as a vector associated with a separation of two identical point charges. Its magnitude is defined as the (positive) charge times the displacement vector between the positive and the negative charge and its direction is from the negative to the positive p = q ■ d, (4.27)

where q is the charge (in Coulombs) and d the displacement vector pointing from - to +. Units of "Debyes" are customarily used where 1D = 3.338x 10~30 Coulomb ■ meters. For N dipoles, in volume V, we define the total electric polarization vector P as the total electric dipole moment per unit volume

The (time-invariant) displacement electric field D in isotropic media is defined as

where e0 and e are the permittivities of free space and sample, respectively, and E is the external electric field and thus

where we have defined k = e/e0 the optical frequency dielectric constant of the material, which is related to the refractive index n via k = n2 . (4.31)

The dielectric permittivity e of a substance is a measure of its ability to "neutralize" part of a static electric field by responding to it with a displacement of some of its localized charge. This charge displacement is referred to as polarization and is not dependent on a material having excess charge. Even for a static electric field, but most importantly when the incident field is time varying, the dielectric permittivity will also depend on time. Because the capacitance (ability to store charge for a given potential difference) C of a medium is directly proportional to its e (as in the elementary case of the parallel plate capacitor where C = eA/d with A the area and d the separation of the plates in the limit d2 ^ A), e can be measured by inserting the medium between the plates of a capacitor and noting the ratio of the capacitance with (C) and without (C0)

the medium so that e = C/C0 . This general basic principle holds even for a fluctuating field but with certain modifications, as will be illustrated later.

Molecular electric polarizability a is a scalar of proportionality that quantifies the polarization of a sample as a result of application of an electric field that in general can have four components: electronic ae (sensitive even to high-frequency fields), ionic or atomic a; (medium frequency), orientational or dipolar ad (low frequency) and interfacial adc (very low to dc frequencies). For simplicity, we will assume that a is an isotropic characteristic of a protein solution sample, which is justifiable at low concentrations. The total polariz-ability as a function of frequency a(w) = ae(w) + a;(w) + ad(w) + adc(w) is a good parameter to use when describing a system such as a protein in solution since, unlike the total dipole moment it does not change as a result of solvation, changes in pH, or local electric field (Eioc) amplitude or direction. We define the total dipole moment as the sum of the permanent dipole moment added to the polarizability-dependent dipole moment p = pperm + aEioc . (4.32)

It can be shown [49] that a molecule in a spherical cavity surrounded by a medium of volume polarization P will experience a local electric field

The above is known as the Lorentz field approximation and is applicable in the case of simple dipolar rotor molecules. Combining (4.28), (4.30), and (4.33) one finds that the average electric dipole moment is a (k + 2\ „

where k = e = relative dielectric constant, and one arrives at the Clausius-Mossotti relation for electric polarizability k — 1N ( 3eny

Was this article helpful?

0 0

Post a comment