Model For The Ageimpaired Medial Temporal System

The Morris water maze is a hippocampus-dependent task that has been described in detail in numerous publications (e.g., Morris, 1984; Gallagher et al., 1993; Bizon et al., 2001). Rodents use visual cues around a circular pool in order to learn the location of a platform submerged 2 cm below the surface of the water's surface (see Figure 32.2).

In our training procedure, rats receive three trials each day for eight consecutive days. All cognitive testing occurs during the light cycle. At each training trial, rats are placed in the water and permitted to swim until they locate the platform, or for 90 seconds, at which time they are placed on the platform. Animals remain there for 30 seconds, and then they are removed from the maze area. The next trial begins again after a 60-second interval.

For spatial learning assessment, the platform stays in one maze quadrant with a randomized starting position for each trial. On every sixth trial, the platform is retracted to the bottom of the pool (probe trial) for the first 30 seconds of the 90-second trial. Our primary measures of hippocampal function are obtained by proximity measures both on the training trials and interpolated probe trials described in detail below. Very importantly, on the ninth day of training, we test sensorimotor skills and escape motivation independent of spatial learning by giving the rats a training session with six trials of cued training. During cued training trials, rats are trained to escape to a visible platform that protrudes above the water and that varies in position on each trial. On each trial, a rat is started in a novel location and has 30 seconds to reach the platform at which point it remains there briefly prior to a 30-second intertrial interval. Cue training is an extremely important control to conduct when characterizing aged rats. It can identify those subjects that are ill (and perhaps not motivated to swim), that have physical impairments that hinder swimming ability, or that have visual acuity problems. Each of these concerns can be alleviated with a proper cue training session incorporated into the water maze protocol. Any rat that fails cue training (i.e., performs worse than the young rat distribution of scores) is eliminated from further experimentation in our model.

Accuracy of performance in the water maze is assessed using a cumulative search error calculated from training

Difference Rat Human Hippocampus

Figure 32.1 Schematic of the medial temporal lobe system, including the hippocampus and associated structures in human and rat brain. The top left diagram shows a mid-sagittal view of the human brain. The general area of the medial temporal lobe is identified and the hippocampal formation is shaded dark gray. The top right panel shows a circuit diagram, and demonstrates the flow of information within this system. Structures that are part of the medial temporal lobe in this system are seen in light gray boxes. Generally, the neocortex sends afferent projections to the hippocampus from the entorhinal cortex via parahippocampal structures. Other subcortical structures (e.g., cholinergic projections from the basal forebrain) also project to hippocampus and entorhinal cortex. The lower diagram shows the rat brain, with the hippocampus highlighted in dark gray. An enlarged coronal section through hippocampus illustrates the basic cellular circuitry within this region. Projections from entorhinal cortex (perforant path) carry information to the granule cell layer (thick black line in C-shape). The hilus is the region encompassed on either side by the granule cell layer of the dentate gyrus. General information flow in the hippocampus entails granule cell projections to the pyramidal neurons of hippocampus proper (i.e., CA3 and CA1, in turn), before exiting the structure via the subiculum and sending information back to the neocortex.

Figure 32.1 Schematic of the medial temporal lobe system, including the hippocampus and associated structures in human and rat brain. The top left diagram shows a mid-sagittal view of the human brain. The general area of the medial temporal lobe is identified and the hippocampal formation is shaded dark gray. The top right panel shows a circuit diagram, and demonstrates the flow of information within this system. Structures that are part of the medial temporal lobe in this system are seen in light gray boxes. Generally, the neocortex sends afferent projections to the hippocampus from the entorhinal cortex via parahippocampal structures. Other subcortical structures (e.g., cholinergic projections from the basal forebrain) also project to hippocampus and entorhinal cortex. The lower diagram shows the rat brain, with the hippocampus highlighted in dark gray. An enlarged coronal section through hippocampus illustrates the basic cellular circuitry within this region. Projections from entorhinal cortex (perforant path) carry information to the granule cell layer (thick black line in C-shape). The hilus is the region encompassed on either side by the granule cell layer of the dentate gyrus. General information flow in the hippocampus entails granule cell projections to the pyramidal neurons of hippocampus proper (i.e., CA3 and CA1, in turn), before exiting the structure via the subiculum and sending information back to the neocortex.

trials and a learning index score calculated from probe trials. This method was originally described and is detailed in Gallagher et al., 1993. Specifically, these proximity measures take into account deviation from an optimal search (i.e., a search that leads the rat directly to the platform). Using computer tracking (HVS Image, UK), the rat's location relative to the platform is sampled in the maze 10 times per second, which gives

Figure 32.2 This figure shows a photograph of a young male Long-Evans rat in the Morris water maze standing on a submerged atlantis platform (HVS Image, UK). Distal visual cues, used to locate the hidden platform, are placed around the periphery of the pool (white geometric figures). Performance in this task is a well-established indicator of medial temporal lobe and, specifically, hippocampal function.

a measurement of the rat's distance from the platform in one second averages. For both probe and training trials, a correction is made for the optimal path based on any given start position, to alleviate bias on trial performance created by variations in the distance to the platform from the different start locations. After this correction is made, the cumulative distance for training trial performance can be summed across trials (see large panel on Figure 32.3) and the average distance from the goal calculated on probe trials (see inset panel of individual Learning Index Scores on Figure 32.3).

Search error has some advantages over traditional measures of water maze like latency (i.e., time to reach platform) or path length (i.e., distance swum to reach platform), particularly for aged rats. By considering the proximity of the rat during the entire trial, one can assume that the most accurate measure of spatial bias is being achieved. For example, one rat may be swimming in close proximity to the platform for 25 seconds, while another rat may be swimming randomly throughout the maze for 25 sec. Although the first rat in the aforementioned example apparently had some appreciation of the platform's location, neither a latency nor path length measure would detect it.

It is also noteworthy that probe trials are critical for the spatial learning assessment of aged rats. Aged rats have a tendency to develop a strategy of ''circling'' the pool at a given distance from the wall, which might result in short latencies and path lengths that are based entirely upon a nonhippocampal dependent strategy. Probe trials are therefore a better means by which to test rats' spatial knowledge, as rats with a good memory of the spatial location will consistently search in the area where the platform had been for at least 30 sec. Probe trials longer than this included throughout the protocol could lead to some extinction of learning. To minimize this problem, we employ an HVS atlantis platform (HVS Image, UK) that can be lowered to the bottom of the tank for the first 30 seconds of the probe trial and then raised again for the last 60 seconds of the trial. We find that this approach minimizes extinction that could be a concern with interpolated probe trials.

The data from these four interpolated probe trials are our main source of individual scores of spatial learning ability. The spatial learning index score is calculated from performance on all four of the interpolated probe trials. Performance on these trials is summed and weighted such that earlier trials receive greater weight (as age differences are generally larger earlier in training). As such, the Spatial Learning Index Score is an individual measure of spatial learning performance that takes acquisition and total spatial performance into account, and provides a single number that can be used to correlate with individual neurobiological data.

Figure 32.3 shows representative data from animals trained in this protocol. Note that as a whole, aged male

Figure 32.3 This graph shows typical performance in the Morris water maze of young (6-month-old) and aged (25-month-old) male Long-Evans rats. Note that both young (open circles) and aged rats (closed circles) performed comparably on the very first trial, demonstrating similar sensorimotor and motivational abilities. As shown in the large panel, both groups learn over the course of time, but the aged rats were significantly impaired in this task in comparison to young adult rats. However, the inset graph in the upper right corner shows that despite an overall impairment in the aged group compared to young, there was substantial individual variability among aged rats such that some aged rats performed on par with young rats and others performed outside the range of young, demonstrating impairment on this task. Note that a higher Learning Index Score indicates poorer performance. About half of the aged rats (A) performed outside the range of the young rats (Y) and were considered aged-impaired, and about half of the aged rats performed as well as the young rats and were considered aged-unimpaired.

Figure 32.3 This graph shows typical performance in the Morris water maze of young (6-month-old) and aged (25-month-old) male Long-Evans rats. Note that both young (open circles) and aged rats (closed circles) performed comparably on the very first trial, demonstrating similar sensorimotor and motivational abilities. As shown in the large panel, both groups learn over the course of time, but the aged rats were significantly impaired in this task in comparison to young adult rats. However, the inset graph in the upper right corner shows that despite an overall impairment in the aged group compared to young, there was substantial individual variability among aged rats such that some aged rats performed on par with young rats and others performed outside the range of young, demonstrating impairment on this task. Note that a higher Learning Index Score indicates poorer performance. About half of the aged rats (A) performed outside the range of the young rats (Y) and were considered aged-impaired, and about half of the aged rats performed as well as the young rats and were considered aged-unimpaired.

Long-Evans rats are consistently impaired relative to young subjects; however, when individual performance (spatial learning index) is calculated, notice that there is much greater variability among aged rats than in young. About half of the aged rats perform on a par with young, and about half perform outside the distribution of young subjects. Assuming that these latter animals are healthy and perform comparably to young rats on cue training, we may interpret that this latter subset of rats is spatially impaired.

MEDIAL TEMPORAL LOBE-BASED COGNITION: CORRELATIONS WITH NEUROBIOLOGICAL CHANGES

Using the derived spatial Learning Index Score, one is then able to correlate a variety of neurobiological and neuroanatomical data with spatial learning performance among aged rats. Figure 32.4 shows data obtained using the behavioral model described above. In this study, the total number of neurons in the hippocampal region was assessed in behaviorally characterized rats using quantitative unbiased stereology, as many theories of age-related memory loss at the time regarded cell loss as the endpoint for age effects on cognition. However, as shown in Figure 32.4, Rapp and Gallagher (1996) found that frank neural loss in the hippocampal formation is not pervasive in the Long-Evans rat model, even though a considerable number of rats included in that study were spatially impaired.

This finding now has been documented for the principal neurons of the hippocampal formation, the entorhinal cortex, and the parahippocampal region, including the perirhinal and postrhinal cortices (Rapp et al., 1996; Rapp et al., 2002; Figure 32.1). The lack of frank neurodegeneration in normal aging has also been confirmed in other rodent models and primates, including humans (Merrill et al., 2003; Rasmussen et al., 1996; West et al., 1994). Additional data supporting the lack of pervasive neurodegeneration in this model are provided by the absence of hypertrophied astrocytes (personal observation; Bizon and Rapp) and other markers such as OX-6 (Nicolle et al., 2001), which are generally associated with widespread degeneration. These data challenge the long-standing view that neurodegeneration is a necessary condition of age-related cognitive impairment, and warrant further consideration.

Certainly, this absence of frank neuronal death has redirected our thinking about age-related neural impairment and focused our efforts toward identifying changes

Figure 32.4 These graphs demonstrate a stereological analysis of neuronal number in all subfields of dentate gyrus and hippocampus proper in young and aged Long-Evans rats previously characterized on the Morris water maze task. Note in panels A and B that the total neuron number is not significantly different between young, aged-unimpaired, and aged-impaired animals for the granule cell layer (A), CA3/2 and CA1 (B). Moreover, no correlation between individual learning index (higher scores indicate greater impairment) and number of neurons in CA3/2 is observed (lower graph). This finding was consistent in both young (open circle) and aged (closed square) animals in hippocampus proper shown here and for the granule cells of the dentate gyrus (data not shown).

Figure 32.4 These graphs demonstrate a stereological analysis of neuronal number in all subfields of dentate gyrus and hippocampus proper in young and aged Long-Evans rats previously characterized on the Morris water maze task. Note in panels A and B that the total neuron number is not significantly different between young, aged-unimpaired, and aged-impaired animals for the granule cell layer (A), CA3/2 and CA1 (B). Moreover, no correlation between individual learning index (higher scores indicate greater impairment) and number of neurons in CA3/2 is observed (lower graph). This finding was consistent in both young (open circle) and aged (closed square) animals in hippocampus proper shown here and for the granule cells of the dentate gyrus (data not shown).

within and communication between neurons that are associated with neuronal plasticity as the main contributors to cognitive deficits. In other words, our current hypotheses are that suboptimal cell functioning and connectivity, rather than frank neuronal death, are likely to underlie the decline of many cognitive abilities that occur with age. Such cellular deficits have in fact been identified in the Long-Evans rat model. For example, differences have been found in the neurophys-iological properties of place cells in aged-impaired rats versus young and aged unimpaired (see Chapter 33 by Nancy L. Nadon). Other findings, including changes in gene and protein expression and deficits in certain cell signaling pathways have been revealed using the Long-Evans rat model.

For example, using young and old Long-Evans rats behaviorally characterized on the Morris water maze as described above, our group has found that metabo-tropic glutamate receptor (mGluRs) and muscarinic cholinergic receptor signaling in the hippocampus is dysfunctional in the aged, learning-impaired rat (Nicolle et al., 1999; Chouinard et al., 1995). Both mGluR

and muscarinic M1 receptor stimulation of phospho-inositide (PI) turnover is blunted in aged rats, and the magnitude of the decrease in signaling correlates with the severity of cognitive impairment. Figure 32.5 shows a correlation between the Learning Index and the PI turnover response that occurred as a consequence of stimulating Type 1 mGluR receptors with the agonist 1S,3R ACPD.

The magnitude of this deficit in PI turnover significantly correlated with the decline in age-related spatial memory (R = —.67, p < .01 when young and aged grouped together; R = —.53, p < .05 when the aged are considered alone). This study illustrates the general approach and power of using naturally occurring rat models to identify age-related neurobiological changes that could underlie memory impairment in the medial temporal lobe system. Further work in this rat model and in aged primates should help confirm that this signaling pathway is a contributor to age-related mnemonic impairment and whether therapeutic interventions directed at this pathway would be beneficial in the reversal of such deficits.

Figure 32.5 This graph shows the relationship of mGluR-mediated phosphoinositide (PI) turnover in the hippocampus in relation to spatial learning ability. Data points represent individual values for young and aged rats. The Learning Index is on the X-axis and the 1S,3R ACPD EMAX, a measure of PI turnover, on the Y-axis. The linear regression between PI turnover and the Learning Index for young and aged rats grouped together can be seen (solid line). A significant negative correlation was observed (R = —0.67, p < .01), with more impaired learning (higher learning index value) related to decreased turnover (lower disintegrations per minute [dpm] of [3H]-IP1). Overall, mGluR-mediated PI turnover decreased with increasing chronological age and was most blunted in the aged rats with the most severe cognitive impairment, indicated by a significant correlation in the aged group alone

Human Brain

Striatum

Prefrontal cortex

Striatum

Prefrontal cortex

Rat Brain

Prefrontal cortex

Figure 32.5 This graph shows the relationship of mGluR-mediated phosphoinositide (PI) turnover in the hippocampus in relation to spatial learning ability. Data points represent individual values for young and aged rats. The Learning Index is on the X-axis and the 1S,3R ACPD EMAX, a measure of PI turnover, on the Y-axis. The linear regression between PI turnover and the Learning Index for young and aged rats grouped together can be seen (solid line). A significant negative correlation was observed (R = —0.67, p < .01), with more impaired learning (higher learning index value) related to decreased turnover (lower disintegrations per minute [dpm] of [3H]-IP1). Overall, mGluR-mediated PI turnover decreased with increasing chronological age and was most blunted in the aged rats with the most severe cognitive impairment, indicated by a significant correlation in the aged group alone

Aging and the Frontal Cortical-Striatal System

Was this article helpful?

0 0
Blood Pressure Health

Blood Pressure Health

Your heart pumps blood throughout your body using a network of tubing called arteries and capillaries which return the blood back to your heart via your veins. Blood pressure is the force of the blood pushing against the walls of your arteries as your heart beats.Learn more...

Get My Free Ebook


Post a comment