Research reportMongolian gerbils learn to navigate in complex virtual spaces
Introduction
Rodents are the most widely used model animals for studying spatial learning and navigation [1], [2], [3] and the underlying neuronal processes [4], [5], [6], [7], [8], [9]. Traditionally, spatial behavior has been investigated in these studies using linear tracks [10] or open fields in various enclosures [11]. Building mazes for more complex spatial tasks [8], [12] is possible with much greater effort but does not overcome the restrictions of the typical lab scale of a few meters. More recently, virtual reality (VR) paradigms have been developed [13], [14], [15], [16], [17], [18], which not only make it feasible to investigate behaviors on arbitrary spatial scales but are also suitable for closed-loop manipulations of the environments, and even allow one to generate physically impossible environments to discriminate between alternative navigation strategies [for reviews see [19], [20]]. Such VR setups allow one to use stable head-fixed preparations in awake behaving animals and combine navigational experiments with advanced recording techniques, such as intracellular recordings [14], [15] or two-photon imaging [21], [22]. In spite of the great success of these VR setups, there were only few attempts to train animals to carry out more complex navigational tasks in virtual environments [23], [24]. Behavioral paradigms in virtual realities for rodents mostly made use of open fields [13], linear tracks [14], [15], [17], [25] or were limited to providing optical flow [16], [22] and spatially unrelated visual stimuli [26]. Moreover, subtle differences in the neuronal space codes have been reported between running in real worlds and VR behavior [18] such that it is not entirely resolved, whether the behavior observed in VR is based on the spatial strategies to a similar extent than in real worlds or whether it is more strongly reflecting direct sensory (visual) stimulation.
In this paper, we report on virtual spatial behaviors in more complex environments, in which the animals are required to perform navigational or behavioral tasks. Since VR setups mainly stimulate the visual modality, we used Mongolian gerbils (Meriones unguiculatus) whose visual system is superior to those of mice or rats [27], [28], [29]. Moreover, spatial navigation in gerbils has been well documented in studies on path integration [1], [30], [31]. Our results demonstrate that gerbils which learned to operate a virtual linear maze were able to make use of their acquired skills in more complex virtual environments. There, the animals exhibited exploratory behavior, even upon first exposure.
Section snippets
Animals
Experiments were performed on adult Mongolian gerbils (Meriones unguiculatus). We used a total of ten gerbils of both sexes. Training started at ages between three and seven months and the animals weighed between 70 and 100 g. The animals received a diet which kept their weight at about 85–90% of their free feeding weight. All experiments were approved according to German Animal Welfare Act and linked European regulations (Reg. von Oberbayern, AZ 55.2-1-54-2532-10-11).
Experimental apparatus
As in previous approaches
Linear maze training
In a first step, we taught animals to move in a linear virtual maze and to orient themselves along straight virtual walls (see Fig. 1C and Section 2). The virtual track was two or four meters long, 23 cm wide and consisted of walls with different textures.
In the first training sessions we did not close the projection screen, i.e., the screen surrounded the animal only to 270°, which gave us easier access to the animal such that we could provide the gerbil with occasional manual feedback to
Discussion
This study demonstrates that rodents can navigate in complex virtual mazes. The stimulated modality was mostly vision and, to a lesser extent, proprioception and the vestibular sense. Our data show that, after training on a virtual linear track for about five days, gerbils accept visual walls without haptic feedback and generalize to more complex virtual mazes. The results support the hypothesis that in VR the animals use spatial strategies to collect food rewards and that they do not simply
Acknowledgements
The authors thank Hansjürgen Dahmen for extraordinary help during the construction of the VR setup, Moritz Dittmeyer for providing the photos and schematic drawing of the setup, and Michael Pecka for helpful comments on the manuscript. This work was funded by the German Ministry for Education and Research (BMBF) via Grant Number 01GQ0440 (Bernstein Center for Computational Neuroscience Munich).
References (35)
- et al.
The hippocampus as a spatial map. Preliminary evidence from unit activity in the freely-moving rat
Brain Res
(1971) - et al.
Temporal encoding of place sequences by hippocampal cell assemblies
Neuron
(2006) - et al.
Hippocampal replay of extended experience
Neuron
(2009) - et al.
A manifold of spatial maps in the brain
Trends Cogn Sci
(2010) - et al.
Progressive transformation of hippocampal neuronal representations in “morphed” environments
Neuron
(2005) - et al.
Hippocampal replay is not a simple function of experience
Neuron
(2010) - et al.
Imaging large-scale neural activity with cellular resolution in awake, mobile mice
Neuron
(2007) - et al.
Sensorimotor mismatch signals in primary visual cortex of the behaving mouse
Neuron
(2012) - et al.
Modulation of visual responses by behavioral state in mouse visual cortex
Neuron
(2010) New methods for analysis of vision in the gerbil
Behav Brain Res
(1981)
Grating acuity of the Mongolian gerbil (Meriones unguiculatus)
Behav Brain Res
Assessing spatial vision – automated measurement of the contrast-sensitivity function in the hooded rat
J Neurosci Methods
Homing by path integration in a mammal
Naturwissenschaften
Place navigation impaired in rats with hippocampal lesions
Nature
Path integration in mammals
Hippocampus
Dynamics of the hippocampal ensemble code for space
Science
Spatial representation in the entorhinal cortex
Science
Cited by (12)
Graded remapping of hippocampal ensembles under sensory conflicts
2021, Cell ReportsCitation Excerpt :The distribution of isolation distances (computed in MATLAB) for all identified place cells was similar to previously published gerbil experiments (Mankin et al., 2019). For a full description, see (Thurley et al., 2014). The virtual reality system was designed to allow animals to freely rotate 360° around their vertical body axis (Hölscher et al., 2005; Thurley and Ayaz, 2017).
A virtual reality time reproduction task for rodents
2022, Frontiers in Behavioral NeuroscienceModality-specific Subpopulations of Place Fields Coexist in the Hippocampus
2019, Cerebral CortexVirtual reality system for freely-moving rodents
2017, bioRxivSpatial cognition in a virtual reality home-cage extension for freely moving rodents
2017, Journal of Neurophysiology
- 1
Contributed equally, in alphabetical order.