Spatial cognition is the ability of an animal to keep track of its location in space by recalling where it has been, which is known as spatial memory (1, 2). Spatial cognition serves as the foundation for spatial navigation, which is the ability to move appropriately and purposefully through the environment (3, 4). Recent studies have demonstrated that the hippocampus plays a critical role in spatial memory consolidation and transitory storage (5, 6), and the peripheral vestibular organs are highly connected via several pathways including the thalamocortical and cerebellocortical pathways mediating head direction information (7–9). Therefore, the vestibular system is suggested to play a critical role in maintaining accurate spatial awareness (2, 4). Unlike other sensory cortices, the vestibular cortex consists of a network of several distinct and separate areas centered in the insular-opercular region (2, 4, 7). In addition, while other sensory systems are organized linearly, vestibular information from peripheral organs becomes multisensory, highly convergent, and highly multimodal when entering the central nervous system (10–13). Canal/otolith interactions in the brainstem and cerebellum at the first synapse, followed by visual-vestibular and proprioceptive-vestibular interactions throughout the central vestibular pathways, enable other sensory and motor signals to be integrated early with vestibular inflow (10, 14).

Numerous animal and human studies have demonstrated that bilateral vestibular loss (BVL) induces a prominent and long-lasting spatial memory deficit by disturbing vestibular-hippocampal interactions (3, 8, 15, 16). BVL results in bilateral hippocampus atrophy, which is associated with spatial memory deficits (17). Though to a lesser extent, unilateral vestibular deafferentation (UVD) has been shown to also impair spatial cognition (18–22). In addition, UVD in rodents has been revealed to lead to long-term changes in the neurochemical and electrophysiological properties of the hippocampus that contribute to spatial cognitive impairments (23–25). Galvanic vestibular stimulation (GVS) exhibits beneficial effects on spatial memory and navigation tasks by modulating the regularity of the firing rate of vestibular afferents (26–28). Further, improvement might in part be due to frequent activation of the vestibular hair cells and the neurons in the vestibular nuclei as well as the hippocampus (22, 29).

Although the peripheral and central vestibular systems are bilateral with ascending pathways on both sides, the multisensory cortical networks in each hemisphere are organized asymmetrically (30). In humans, vestibular information processing shows a hemispherical preference with a dominance determined by handedness (within the right hemisphere in right-handers and the left hemisphere in left-handers) (31–37). The right hemispheric preference for vestibular signal processing was observed in functional and structural imaging analyses showing stronger connectivity values, larger anatomical nodes, and higher functional vulnerability in the right hemisphere in right-handers (38). These findings are consistent with a well-documented superiority of the right hemisphere for visuospatial tasks and navigation (37, 39). In addition, the vestibular input to the ipsilateral hemisphere has a significant preponderance compared with the contralateral hemisphere (33, 37, 40). This lateralization of vestibular information processing was observed in ontogenetic older species. Vestibular processing in rodents was reported to be dominated by the left hemisphere regardless of handedness (41, 42). In rodents, the left-sided vestibular information is processed by more complex cortical and subcortical networks than the processing of information from the right-sided vestibular input (41). Considering that acute lesions in the dominant hemisphere can result in more severe symptoms, such as aphasia or neglect, than lesions in the non-dominant hemisphere do (43), it was speculated that the cognitive outcomes of left-sided or right-sided UVD can be different. However, only a few studies have investigated spatial cognition concerning the side of vestibular loss (44).

In the current study, we analyzed the differential effects of acute right- and left-sided UVD on the higher vestibular spatial cognition and locomotion reflecting hemispheric dominance using a mouse model with surgical labyrinthectomy. We also evaluated whether there was a differential effect of GVS on the deficits depending on the affected side.

During the acute phase after surgery, signs of UVD, such as spontaneous horizontal nystagmus beating toward the contralesional side, head-tilting, falling toward the ipsilesional side, disturbance in backward gait, and clockwise circling, were observed in both Lt.-UL and Rt.-UL groups. The control animals that underwent sham operations did not show these symptoms. It took ~2 days after UL for the mice to regain a stable posture and walk steadily. Based on this observation, we conducted all behavioral investigations starting from POD 3 when the mice were free from the limitations of motor coordination problems (Figure 1).

Vestibular information has been demonstrated to have a major role in determining egocentric heading and rotations around an earth-vertical axis for accurate spatial performance (74–77). Given the critical function of the vestibular system in spatial orientation with three-dimensional coordinates (17, 30, 31, 78–80) via its multisensory, highly convergent, and highly multimodal mechanisms (10–13, 31), it is reasonable to speculate that unilateral vestibular lesions will affect spatial cognition, as shown in our previous study (22). Both right and left UVD impaired short-term spatial working memory and spatial reference memory during the week after UL, but the deficits recovered within 2 weeks. Long-term spatial cognition, as evidenced by the longer escape latency in hidden platform trials and a decrease in the percentage of time spent in the target quadrant in the probe trial at 2 weeks after UL, was impaired in the UVD groups compared with the control group. Although some processes are shared in the hippocampal CA1 subregion, the short-term and different phases of long-term memory are not sequentially linked, and the consolidation of new memory into long-term memory is time-dependent (66, 81–83). While short-term memory is formed almost immediately and is disrupted by subsequent learning, long-term memory requires consolidation over time as a result of hormonal and neurological influences on memory and the involvement of molecular and cellular mechanisms (66).

Our findings corroborated previous research indicating that acute UVD can cause spatial cognition deficiencies, especially transient short-term and more lasting long-term memory cognition deficits (18–20, 22, 84). Multiple pieces of evidence support our findings, including studies of electrical changes [electrical excitability (85–87), particularly long-term potentiation (86, 88)] and biochemical changes [neuronal nitric oxide synthase expression (23, 85, 89–91), N-methyl-D-aspartate receptor subunit expression (24), glucose metabolism (92), and glucocorticoid receptor expression (93)] and cellular proliferation [Arc, zif268, c-fos gene expression (51, 94, 95)] in the brain, in particular in the hippocampal formation, of UVD individuals both in vitro and in vivo. Several studies have revealed that these alterations can persist for an extended duration following UVD lesions, for 1 month (86) or 5–6 months (85), although the studies did not differentiate between short-term and long-term spatial memory deficits.

The vestibular system integrates multisensory signals, most notably vestibular and visual input (96), between the ipsilateral and contralateral sides of the multi-level brain regions. Despite the lack of animal data for comparison of behaviors between left- and right-sided UL animals, some results from human studies explain the current experimental findings as vestibular lateralization differences reflect evolutionary development between the two species (31, 41). The prevalence of hemispatial neglect and the frequency of spatial processing deficits were significantly higher in right-handers with right hemispheric strokes compared with those of left hemispheric strokes (97). In parallel, pusher syndrome, in which patients actively push away from the non-hemiparetic side and show postural imbalance (98), is observed with a significantly higher frequency in patients with right hemispheric stroke than in those with left hemispheric stroke, and slower recovery from pushing is shown after right hemispheric stroke (99). In addition, cortical and subcortical activation by vestibular caloric stimulation depends on the handedness of the individuals and the side of the stimulated ear, i.e., activation was bilateral but predominant in the hemisphere ipsilateral to the stimulated ear and exhibited right-hemispheric dominance in right-handers or left-hemispheric dominance in left-handers (31). With regard to the pathological state, a recent (18). F-fluorodeoxyglucose (F-FDG) PET study reported that brain activity in the acute phase of right- and left-sided UVD exhibits different compensatory patterns, in which the dominant ascending input is shifted from the ipsilateral to contralateral pathways, presumably due to the missing ipsilateral vestibular input (100). This might imply that the vestibular “dominant” right ear lesion might show more severe consequences than the vestibular “non-dominant” left ear lesion in humans (100, 101). Another study using H₂O¹⁵-PET imaging demonstrated that the side-specific suppression of vestibular cortex activations was more pronounced in patients with right-sided thalamic lesions than those with left-sided lesions (37). These data demonstrated the functional importance of the dominance of ipsilateral vestibular ascending pathways from the end-organ (ear) to the relay-station (thalamus) and to the vestibular dominant right hemisphere in right-handed individuals (33, 37, 40).

Several human studies have investigated spatial cognition concerning the side of vestibular loss; however, their findings remain controversial (20, 44, 84, 100–102). One study demonstrated differential impairment of embodied spatial cognition between left- and right-sided UVD patients, with left-sided UVD patients more severely affected (44). However, other studies showed that the spatial memory and navigation of right-sided UVD patients were more severely affected (101, 103). The right-sided UVD patients performed significantly worse in the probe trial of virtual MWM, i.e., spending less time and distance searching in the correct quadrant and having a higher heading error than the left-sided UVD patients or controls (101). These findings are compatible with the recently described dominance of the right labyrinth and the vestibular cortex in the right hemisphere (37, 100, 101). This discrepancy can be explained by the variety of spatial cognitive tasks in the study that were driven by distinct brain areas that are lateralized at various levels (44).

Vestibular lateralization studies in animal models such as rodents have been promoted. A recent microPET study revealed that the vestibular processing in rats follows a strong left hemispheric dominance independent from the “handedness” of the animals (41). The authors showed that the left vestibular information was processed by a complex cortical and subcortical network, whereas the right vestibular input was processed by fewer cortical areas, which suggested dominant left-sided vestibular information processing (41). Phylogenetic concepts have led to the hypothesis that the difference in vestibular dominance between rodents and humans is the result of the evolution of speech and handedness in humans, both of which represent mechanisms that could have led to reconfiguration within the vestibular cortical network (31, 41). Considering the vestibular lateralization of the left hemisphere in rodents (41), we hypothesize that the missing dominant ipsilateral input resulted in a more serious outcome than those occurring in missing input of the non-dominant side.

In our previous study, we allocated the cathode to the lesion side (right) and the anode to the intact side (left) for GVS, which resulted in a significant improvement in spatial cognition as well as locomotion and VOR in the UL mouse model (22, 49). In the current study, we used the same protocol for GVS (CLAR in the Lt.-UL and CRAL in the Rt.-UL groups), which helped to rebalance the firing rate with attenuating the intact side and facilitating the lesioned side (26, 28) and accelerate vestibular recovery. Although this study corroborated previous findings of the effects of GVS in improving UVD-induced spatial cognition impairments (22, 104), the precise mechanism remains unknown. However, increasing evidence has shown that GVS enhances the function of spatial navigation through multimodal mechanisms (22). The GVS currently likely operates by modulating the firing rate of vestibular afferents (26, 28), and GVS also excites medial vestibular nuclei and increases the firing rates of hippocampal CA1 complex spike cells corresponding to place cells (105). The electrical stimulation of afferent vestibular fibers evidently enhances the long-term potentiation and long-term depression in the vestibular nuclei of rats in vitro (106–109), which in turn facilitates the effects in the hippocampus. Similarly, GVS has been shown to generate theta activity in numerous areas of the hippocampal formation (110), which plays a pivotal role in spatial information processing and modulates self-movement signals (111). This formation also improves neuronal activity for spatial orientation (112, 113). Additionally, an increase of c-Fos-positive cells in the hippocampus, which is an indicator of neuronal activation, was detected following subsequent repetition of GVS (29, 95).

Our findings showed, for the first time, the differential effects of GVS intervention depending on the lesion sites. The Lt.-UL mice group showed greater improvements after GVS intervention in both short- and long-term spatial memory, which might be due to poor performance in the Lt.-UL mice compared with the Rt.-UL mice before GVS application. Alternatively, this selective effect was due to the results of a functional asymmetry of the vestibular apparatus and a set of common neural mechanisms between each GVS paradigm (CRAL vs. CLAR) and brain regions (114). In a study with healthy subjects with functional MRI (fMRI), the CLAR activated both hemispheres, whereas the CRAL activated only the right hemisphere (115). However, there might be differences between the GVS protocols as well as between patients with UVD and normal individuals; (116) this should be examined in future studies.

In conclusion, our results showed that right- and left-sided UVD impairs spatial cognition and locomotion differently and exhibits different compensatory patterns in mice. Considering that the vestibular processing in rodents follows a strong left hemispheric dominance, ipsilateral peripheral vestibular injuries on the dominant side have more severe consequences and slower recoveries than those on the non-dominant side. We also identified the effects of bipolar GVS on accelerating recovery for UVD-induced spatial cognition when the cathode (stimulating) was assigned to the lesion side. The current findings suggest that sequential bipolar GVS intervention might have substantial implications for comprehensive and specialized management in patients with impairment of spatial cognition, especially induced by unilateral peripheral vestibular damage on the dominant side.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

The animal study was reviewed and approved by the Animal Care Committee of Gachon University of Medicine and Science (IRB MRI2019-0008).

S-YO involved in study concept and design and drafting a significant portion of the manuscript. TTN, G-SN, and J-JK involved in acquisition and analysis of data and drafting a significant portion of the manuscript and figures. GCH and J-SK involved in acquisition and analysis of data. J-SK and MD involved in analysis of data and drafting a significant portion of the manuscript. All authors contributed to the article and approved the submitted version.

This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (Ministry of Science and ICT) (No. 2019R1A2C1004796), and by the BK21 FOUR Program by Jeonbuk National University Research Grant.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.