Mapping brain activity in vivo during spatial learning in mice
Deficits in the encoding, retrieval and manipulation of sensory or memory information in the brain contribute to a number of neuro- and psychopathologies in humans. To better understand the underlying principles of memory processes, efforts still have to rely on animal research. The laboratory mouse...
|PDF Full Text
No Tags, Be the first to tag this record!
|Deficits in the encoding, retrieval and manipulation of sensory or memory information in the brain contribute to a number of neuro- and psychopathologies in humans. To better understand the underlying principles of memory processes, efforts still have to rely on animal research. The laboratory mouse as a model organism provides many advantages to study the underlying mechanisms as one can directly interfere with brain functions via a number of tools including genetic manipulation. For this reason, it is very important to have memory tasks that match human memory testings and suits the characteristics of mice. One cognitive capability that is highly preserved across species is spatial learning, what enables the daily required relocation of certain positions in space. Mice and men have developed two different strategies to do so, which rely on different types of reference information. During place learning, environmental cues are adopted and incorporated into a cognitive map. In contrast, response learning is based on directed movements along a specific route. Place and response learning can also be dissected in terms of their biological substrate. While place learning requires the hippocampus (HPC), response learning depends on the striatum. Most behavioral tests, that can clearly distinguish between the two strategies were originally designed for rats. Since the behavior of mice and rats can be considerably different, it is necessary to adapt these tasks to the requirements of mice. In order to improve then the translation of structural and functional results from rodents to humans, methods must be applied that match the coarse in vivo imaging typically used in humans. Manganese-enhanced magnetic resonance imaging (MEMRI) may therefore be useful, as it provides three-dimensional maps of the living mouse brain. Since manganese increases the brain contrast in magnetic resonance (MR) images, MEMRI is regularly applied to depict the volume of specific brain structures in vivo. A second feature of manganese is, that it enters neurons through L-type voltage-dependent calcium channels (LTCCs) and therefore might be an indicator for neuronal action. However, ithas never been shown that LTCCs directly regulate the MEMRI contrast in vivo, which would be essential to establish it as a functional tool in order to measure brain activity in the living mouse brain. The application of MEMRI to cognitive tasks might then be helpful to identify underlying brain circuits and in combination with other techniques also essentially involved neurobiological mechanisms. Finally, it may also increase the comparability of human and rodent research. Therefore, I wanted to establish the water cross maze (WCM) as a suitable tool to study different learning strategies in mice and relate them to HPC functioning. Next, I aimed to dissect the influence of LTCCs on MEMRI contrast (specifically Cav1.2 and Cav1.3 as the two major LTCCs in the brain) in order to justify a functional application of MEMRI. At last, MEMRI should be implemented to depict learning processes in the WCM before I wanted to interfere with LTCC functioning to further explore their role in spatial learning. I had been able to demonstrate that theWCMwas particularly suitable for mice because it prevented most unwanted strategies that mice often adopt during the Morris water maze task. Further the test clearly dissected response from place strategies, which were both successfully acquired by C57Bl/6N mice. However, mice failed to relearn under response training independent of the original navigation strategy that was adopted within the week before. These results suggested, that not only place learning but also relearning is predicated on the HPC. Accordingly, HPC-lesioned mice were unable to acquire a place strategy, however, they adopted a response strategy instead. Further, relearning was blocked by the lesion, if less than 40% of the entire HPC remained. The inability to relearn was best reflected in the residual volume of the left ventral HPC. Second, I investigated the contribution of Cav1.2 and Cav1.3 on MEMRI contrast with the help of corresponding knockout mice. I was able to demonstrate that the Cav1.2 knockout affected at least 50% of the manganese-dependent contrast increase seen in MR images, whereas Cav1.3 knockouts caused no significant alterations. In addition, a locally defined knockout of Cav1.2 induced contrast differences in a projection region far away from the knockout side suggesting a bias in contrast differences away from the soma towards the axon terminals. Overall, this indicated a voltage-dependent manganese displacement in the brain and therefore suggested the functional application of MEMRI. For this reason, I combined place training in the WCM with manganese injections to map brain activity in vivo. On the one hand, the accuracy score was related to a fear associated network comprising the basolateral amygdala (BLA) and ventral HPC. On the other hand, the latency correlated with the dorsal HPC, specifically the left CA3 and the right CA1 region. First, this was in line with functional magnetic resonance imaging (fMRI) results obtained in humans, where the left HPC indicated response navigation and the right place memory formation. Second, the associations indicate the integration of emotional information into cognitive processing. At last, learning and relearning capabilities of Cav1.2 knockout mice were explored. Despite reduced MEMRI intensities in learning associated regions, knockout mice successfully acquire place and response memories and were also capable to revert the place memory afterwards. However, animals exhibited a significant retardation during place learning ,which can be attributed to impairments of late long-term potentiation (LTP) in the CA1 region of the HPC. Overall, the WCM suits the characteristics of mice and allows the distinction of different learning strategies. Further, mice similar to rats require an intact HPC to use place strategies in the WCM and at least 40% of the total HPC volume is necessary to accomplish relearning. Since I could demonstrate for the first time that MEMRI contrast largely depends on Cav1.2, MEMRI was employed to map brain activity in freely moving mice. I could identify brain regions most in the HPC that correlate with place learning parameters in the WCM for the first time in vivo. These results further match findings in humans, where place and response learning occur in parallel during place navigation in the left and right HPC, respectively. In addition, they suggest the integration of emotional information into cognitive precessing. At last, Cav1.2 is involved but not essential for place learning in the WCM. Future investigations with temporary knockouts might be useful to further elaborate the role of Cav1.2 in learning and memory functions.