M.Sc Thesis

M.Sc StudentBorodach Ilia
SubjectContribution of Visual System to the Stability of
Head-Direction Cell Responses
DepartmentDepartment of Medicine
Supervisor ASSOCIATE PROF. Dori Derdikman
Full Thesis textFull thesis text - English Version


The hippocampus and related regions are playing a role in our working memory, but also in storing and processing of the spatial information. They also include different types of neurons which fire in correlation with the location or the orientation of the rat in space. We focused on Head-Directrion (HD) Cells - neurons which fire with correlation of rat's head-direction. These cells are a kind of biological compass, firing without relation to the location in the environment, but only to orientation. It's now known that HD cells exist at least in those regions: post-subiculum, parasubiculum, retrosplanial cortex, the thalamus, the lateral mammilary nucleus, the dorsal-tagmental nucleus, the striatum and the entorhinal-cortex.

A rat pup opens its eyes only at P15. Two recent papers (Bjerknes et al., 2015; Tan et al., 2015) demonstrated that HD cells are active before eye-opening (they are probably influenced by the vestibular system), but they are not stable and tend to drift. We analyzed the data of (Bjerknes et al., 2015) which was recorded in single units, and we quantified the phenomenon related to time. We compared the data to (Sargolini et al., 2006b; Derdikman et al., 2009; Bonnevie et al., 2013), which was recorded in adult rats. We also had the data divided by the arenas’ geometry-shape (circle and square), and we used it to make comparisons between different arena shapes.

            We created a new method: we divided the trajectory time to slices of 100 seconds, with jumps of 10 seconds (there is an overlap between the slices), and we measured the "temporary" head-direction of the cell. Relying on this we measured the drift in the cells which were defined as HD cells. We found that the drift after eye-opening was much smaller than the drift before eye-opening, which means that probably the visual system contributes to the stability of HD cells. We also showed that the HD cells continue to improve and become more stable at adulthood.

            We also found that the exponent of the accumulated error is approximately, which may imply that there is a path-integration process for head-direction signal generation in the rat's brain. We further found that, unlike in pups, the accumulated drift in adult rats was close to 0, demonstrating that the signal becomes very stable at maturity.

            We also used the same method to check the difference between circular and square arenas. We found that in rat pups after eye opening there is a major difference between the arena shapes, such that the drift in squares is clearly smaller than the drift in circles. In adult rats we found the same phenomenon but with much smaller difference. This suggests that the geometry of the arena is used by the HD cells in order to correct for errors in orientation.

            In correspondence with the literature, our results thus suggest that HD cells are generated by path-integration mechanisms.