Spatial orientation and the brain: identifying the link between neural representations of direction and location

The ability to recognise locations and navigate between them is essential for both humans and other mobile animals. A central question in neurobiology is how the mammalian brain achieves this. One of the most tractable ways in which this can be addressed is by recording neurones in specific areas of the rodent brain. As the essential structure of the rodent brain is similar to that of other mammals, including humans, this approach allows identification of the basic mechanisms for spatial orientation. In the current research, we will use neuronal recordings and brain manipulations to see how neural representations of location and direction are linked.

In the rat brain, researcher have discovered place cells, which encode specific locations in an animal's environment, and head direction (HD) cells, which are neurons that are tuned to the specific direction and organism might face. More recent developments include the discovery of grid cells in the rodent medial entorhinal cortex. These cells show repeated, grid-like firing patterns as the animal explores and enclosed environment. An additional type of spatially-tuned neuron is the boundary/border (B/B) cell, which exhibits a firing field a given distance and direction from an environmental boundary. Although much is known about the individual properties of these types of spatial cells, what is not known is how they interact to guide spatial behaviour.

One clue about how cells interact arises from a recent study that we conducted. It was based on findings by Spiers et al. (2015), who had shown that place cells in the rat show similar fields across four, parallel rooms in a maze. This suggests that place cells encode local maps, and that these maps are similar in rooms that look the same. We replicated this, but found that place cells do not show similar fields across maze rooms if the rooms face different directions (Grieves et al., 2015). Together, these results suggest that place cells are driven by room boundaries and by inputs which encode direction.

In the proposed experiments, we will exploit this phenomenon to study how HD cells interact with place cells. Our hypothesis is that HD cells underlie the spatial firing of B/B cells, and that this gives rise to the ability to discriminate similar environments which face different directions. To test this, we will see whether head direction cells show an unchanged preferred firing direction when the animal travels between four identical maze compartments which face different directions. We will also test whether B/B cells are sensitive to the global orientation of local compartments.

In a second series of experiments, we will remove the HD cell system (by selectively removing the lateral mammillary nuclei (LMN), a brain region critical for generating the head direction signal). In rats without the LMN, we predict that B/B cells will no longer discriminate between the direction of local compartments, and will fire in the same way in each compartment.

In a third series of experiments, we will test whether the HD cell system is necessary to tell local compartments apart behaviourally by again removing the LMN. In the fourth series of experiments, we will target the putative HD projections of the LMN (to the anterior dorsal thalamus, another brain region containing head direction cells), to test whether it is these projections specifically that are essential for orthogonal place fields in repeated compartments.

The proposed experiments will address basic questions about how representations of direction and location in the brain interact. Though much is known about the individual properties of place, HD, grid, and B/B cells, what is not known is how these representations work together to guide seamless navigation. The proposed experiments will advance the field by linking these systems, and thereby identifying one mechanism by which animals can distinguish similar environments.

Neural representations of direction and location are evident in the spatially selective firing or head direction (HD) and place cells, respectively. Place cells receive inputs from grid cells, and, putatively, boundary/border (B/B) cells. How these representations relate to one another and how they underlie spatial cognition is poorly understood. We will address this gap in our knowledge by exploiting the phenomenon of place cell repetition.

1) We will test whether the B/B cells, grid cells, and HD cells are sensitive to the orientation of local environments. We have demonstrated such a sensitivity for hippocampal place cells, where place fields repeat in environments that face the same direction, but not for those facing different directions (Grieves et al., 2015). We hypothesise that this sensitivity to direction arises from HD cell inputs to B/B cells.

2) To directly assess they hypothesis that HD cells drive B/B cell spatial firing, we will test whether removal of the HD cell system (via lesions of the lateral mammillary nuclei (LMN), essential for the generation of the HD signal) causes a loss of sensitivity to local environment orientation in B/B cells.

3) Thirdly, we will test the hypothesis that the HD cell system is necessary to distinguish local environments which face different directions. This will be done using a novel odour/location discrimination task that we have developed, which is particularly sensitive to direction.

4) In the final experiment, we will test whether it is the projections from the LMN to the anterior dorsal thalamus (AD) that are involved specifically in disambiguating local environments facing different directions. We predict that when the LMN -> AD connections are inhibited, place cells and B/B cells will be insensitive to the orientation of local environments.

Together, these experiments have the potential to link the HD cell system with representations of location, and with spatial discrimination learning.

Who will benefit from this work?

The proposed work is comprised of basis research on how the neural systems that represent direction interact with the neural systems that represent locations. Potential beneficiaries of this work beyond fellow researchers will likely include individuals with an interest in spatial cognition (e.g., those interested in navigation systems or their design), patients and caregivers of patients with topographical disorientation or Alzheimer's disease (where difficulties in spatial location recognition are common), designers of care homes and students interested in learning and memory.

How will they benefit?

Robust representations of location and direction are found in the spatially-tuned firing of place cells, grid cells, boundary/border cells and head direction cells, but how these different representations fit together to guide spatial cognition is not understood. The proposed studies will attempt to link between these representations, and thereby identify the basic brain circuit underlying the discrimination of local environments.

Understanding the normal representation of orientation may provide the standard from which abnormal brain function can be estimated. To be specific, in pathological conditions such as Alzheimer's disease, loss of neurons in the entorhinal cortex occurs early in condition. This same brain region contains the types of neurons - head direction cells and boundary/border cells - that will be manipulated in the current experiments. Thus, understanding how these neurons function in the intact brain, may allow development of, for example, early diagnostic tests based on an impaired direction sense.

Understanding how neural representations of direction contribute to the ability to discriminate environments with similar features may also have implications for the design of care homes. There is evidence that aged individuals have difficulties navigating within nursing homes (Passini et al., 2000). Our work may indicate that a sense of direction facilitates the discrimination of similar environments when they face different directions. If such a finding reflects a basic principle of spatial cognition, it could be used to improve the design of residential facilities with multiple, similar rooms. To that end, we will organise a set of workshops on neuroscience and design, which will bring together neuroscientists and architects to consider how neural representations of space might be leveraged to enhance design (see Pathway to Impact document).

The proposed work may benefit students interested in learning and memory by providing a template for understanding how other forms of cognition operate. In addition, characterizing the relationship between representations of direction and location could be applied to the design of mobile robots.