The note is an excerpt from the Shinder et al. 2019. We just describe some key conception and results in this study as a significant note. Please read the original paper if you are interested in the study on https://www.physiology.org/doi/full/10.1152/jn.00880.2017 .

Shinder, Michael Elliott, and Jeffrey Steven Taube. “Three-dimensional tuning of head direction cells in rats.” Journal of Neurophysiology 2019 121:1, 4-37

Shinder et al. 2019 monitored head direction cell responses from rats in three dimensions using a series of manipulations that involved yaw, pitch, roll, or a combination of these rotations. Results showed that head direction responses are consistent with the use of two reference frames simultaneously: one defined by the surrounding environment using primarily visual landmarks and a second defined by the earth’s gravity vector.

The head direction system encodes the direction the animal is facing in the environment. A given HD cell fires only when the animal faces a particular direction. When the animal moves and turns its head, thus changing its directional heading, different HD cells become activated and cells that were previously active become deactivated. Thus the HD cells that are active at any given moment are the ones that have a preferred firing direction for the direction that the animal’s head currently faces.

Precisely how the brain processes self-motion signals into a HD signal remains unknown.

One sensory signal of primary importance for maintaining an accurate representation of heading is the vestibular system, which is sensitive to angular and linear acceleration of the head. (Shinder and Taube 2010; Stackman and Taube 1997)

The issue of how the brain represents three-dimensional space has been difficult to resolve as terrestrial animals do not explore space as a volume and current examples of bats that do navigate volumetric space come with caveats of intrinsic differences in the types of neural signals available to encode space (Barry and Doeller 2013)

Two possibility exist for how the HD signal is organized on the plane, as well as ultimately represented in 3D space.

One is that directions are presented in the animal’s plane of locomotion.  This possibility means that the vestibular signal would be used to update a HD signal that is referenced to the orientation of the head relative to the animal’s body.

Another possibility is that the reference frame used by HD cells is fixed to the environment and is independent of the animal’s plane of locomotion.

Finkelstein et al 2015, proposed that HD cell firing in bats was tuned in 3D and could be modelled as a 3D toroid.

More recently, two studies have suggested the importance of gravity in controlling the reference frame for HD cells.

Laurens et al. (2016) monitored cells in the macaque anterior thalamus and identified cells that were tuned to pitch and roll orientation relative to gravity, independent of visual landmarks.

Page et al. 2018 proposed a model of HD cell firing in the vertical plane that stipulated how cells would shift their PFDs when an animal moved around vertically oriented corners. The model proposed that the HD system treated 3D space as two interrelated 2D surfaces and that cell firing was updated based on how the animal rotated its head about two different axes: 1) yaw rotations around the animal’s dorsal-ventral axis, and 2) rotation of the animal’s dorsal-ventral axis with respect to the earth’s gravity axis. This model was referred to as the dual-axis model, and some preliminary data were presented that was consistent with it.

For species that inhabit a more 3D environment, like bats, how a more 3D reference frame is constructed in allocentric coordinates, and whether the process is completely different that the one present in rodents, or just added onto the 2D gravity dependent process, is not known.  Perhaps the 3D coordinate system present in bats, which enables 3D tuned HD cells, has evolved from a gravity-dependent coordinate system into a gravity independent system, which has allowed for a more complete 3D spatial representation. Finally, another issue to consider is whether a gravity-dependent reference system is universal in the brain and used by other spatial cells types, such as place cells and grid cells. While this remains an open issue, the authors presume that the brain has only evolved one coordinate system to represent an animal’s spatial orientation relative to its environment. Thus they suspect that the coordinate frame used by place and grid cells will also be shown to be referenced relative to gravity.

This study was designed to determine whether the reference frame used by HD cells is fixed to the environment and independent of the animal’s plane of locomotion or does it rotate with the animal into different planes as the animal transitions across different surfaces that are orthogonal to one another.

Their results showed that gravity play an important role in defining the HD cell’s reference frame, but based on previous work (Taube et al. 2013), local view cues also play a critical role and can override an earth-bound reference frame.

Thus the reference frame for the HD signal is determined by a combination of the two factors: gravity and environmental landmark cues that can be local or global. Furthermore, the findings do not provide strong support for an internal model that is based on multiple sensory and motor cues to represent directional heading; rather, the internal model appears optimized for inputs from the horizontal canals although it can be modulated by proprioceptive and motor efference cues.

Shinder, Michael Elliott, and Jeffrey Steven Taube. “Three-dimensional tuning of head direction cells in rats.” Journal of Neurophysiology 2019 121:1, 4-37