Differential Impact of Multiple Sensory Deprivation on Spatial-coding Cells in Medial Entorhinal Cortex

This study demonstrates that the Medial Entorhinal Cortex flexibly integrates visual and tactile sensory inputs to sustain spatial coding, revealing that while grid and spatial cells rely primarily on vision, they can effectively anchor to tactile cues in darkness, whereas head direction and border cells are disrupted by the loss of either sensory modality.

The Gist

Imagine your brain is like a GPS system inside a car. To know where you are, this GPS needs two things: a map it built itself (based on how far you've driven and which way you turned) and landmarks outside the window (like a red barn, a stop sign, or a mountain) to correct its position.

This paper is about what happens to that GPS when you cover up the windows (remove vision) and tape over the car's bumpers (remove the sense of touch). The researchers studied the Medial Entorhinal Cortex (MEC), a specific part of the brain that acts as the "map-making engine" for navigation. They used mice, whose whiskers are like super-sensitive touch antennas, to see how the brain builds a map when it can't see and can't feel.

Here is the story of their discovery, broken down simply:

1. The Setup: The "Blindfold and Glove" Experiment

The researchers put tiny cameras inside the brains of mice to watch their neurons fire while they ran around a square room.

• Phase 1 (Normal): The mouse runs with lights on. It sees the walls and feels the floor. The brain builds a perfect map.
• Phase 2 (Darkness): They turn off the lights. The mouse is blind but can still feel the walls with its whiskers.
• Phase 3 (The "Glove"): They trim the mouse's whiskers. Now the mouse is in the dark and can't feel the walls with its whiskers. It's floating in a void.
• Phase 4 (The Twist): In a second experiment, they made the room full of rough sandpaper patches (tactile landmarks) but kept it pitch black.

2. The Characters: Different Types of Brain Cells

The brain has different "employees" working on the map, and they react differently to the sensory deprivation:

• The Compass (Head Direction Cells): These cells tell the mouse which way is "North."
  • The Result: When the lights went out, the compass got a little shaky. When the whiskers were cut, the compass spun wildly. It turns out, the compass needs both sight and touch to stay steady.
• The Boundary Guards (Border Cells): These cells fire only when the mouse is near a wall.
  • The Result: These guys are very sensitive. Whether it's dark or the whiskers are gone, they get confused. They rely heavily on physically bumping into walls or seeing them to know where the edge of the world is.
• The Grid Makers (Grid Cells): These are the stars of the show. They fire in a perfect honeycomb pattern, creating a coordinate system (like graph paper) for the whole room.
  • The Result: In the first experiment (lights on, then dark), the Grid Makers were surprisingly tough. When the lights went out, they struggled a bit, but when the whiskers were cut, they actually held their ground! They seemed to rely mostly on the visual map they had already built.
  • However, in the second experiment (pitch black room with sandpaper landmarks), the Grid Makers needed the whiskers. Without the whiskers to feel the sandpaper, their perfect honeycomb map fell apart.

3. The Big Discovery: The Brain is a Flexible Adapter

The main takeaway is that the brain is like a smartphone with a backup battery.

• When you have Vision: The brain uses sight as the primary battery. If you take away the touch (whiskers), the brain switches to "Power Saving Mode" and relies on the visual map it already made. The map gets a little fuzzy, but it still works.
• When you have NO Vision: The brain switches to "Touch Mode." It uses the whiskers to feel the sandpaper and walls to build the map. In this mode, if you cut the whiskers, the battery dies completely. The map collapses because the brain has no other way to know where it is.

The "Sandpaper" Analogy:
Imagine you are in a pitch-black room. If the floor is smooth, you have no idea where you are. But if you put a piece of sandpaper in the corner, your whiskers can feel it, and your brain says, "Aha! That's the corner!" The brain uses that tactile clue to anchor the whole map. When the researchers cut the whiskers, the mouse lost that anchor, and the map dissolved.

4. The "Whisker-Responsive" Neurons

The researchers also found a special group of neurons that act like motion sensors. These cells light up specifically when the whiskers move, even if the mouse isn't moving its body. It's like having a dedicated team in the brain whose only job is to say, "Hey, the whiskers just brushed against something!" This proves that the brain has a direct line to process touch specifically for navigation.

5. The Final Lesson: The Wall is the Anchor

In the most extreme test, they removed the walls entirely (leaving the mouse on a floating platform). Even with the sandpaper on the floor, the map fell apart.

• Why? Because the brain needs boundaries. Just like a house needs walls to define the rooms, the brain needs the physical sensation of a boundary to know where "inside" ends and "outside" begins. Without walls, the brain's GPS loses its reference point entirely.

Summary

This paper tells us that our internal GPS isn't rigid. It's a chameleon.

• If you can see, it ignores the touch.
• If you can't see, it relies heavily on touch.
• If you take away both, the map breaks.
• And if you take away the walls, the map vanishes completely.

It shows that the brain is incredibly adaptable, constantly swapping which senses it trusts to keep us from getting lost, but it always needs some anchor to the real world to function.

Technical Summary

1. Problem Statement

Spatial navigation relies on the integration of internal self-motion signals (path integration) with external sensory cues (visual, tactile, olfactory) to anchor internal spatial maps. While the role of visual landmarks in calibrating the Medial Entorhinal Cortex (MEC)—the hub for spatially modulated cells like grid, border, and head-direction (HD) cells—is well-established, the specific contribution of tactile sensation (specifically whisker-mediated input in rodents) to spatial coding remains poorly understood.

The study addresses two primary questions:

1. How do visual and tactile deprivation differentially affect the stability and tuning of distinct MEC cell types (HD, border, grid, and non-grid spatial cells)?
2. Can tactile cues alone sustain spatial representations in the absence of vision, and how does the removal of tactile input degrade these representations?

2. Methodology

The researchers employed miniature two-photon calcium imaging to monitor neural activity in the MEC of freely moving male mice expressing GCaMP6 (s or f). The study utilized a multi-stage experimental design across two distinct environments:

• Environment 1 (Visual Cue-Enriched): An open-field arena with three distinct visual cue cards on the walls.
  • Protocol: Mice underwent sequential sessions: Light (L) $\rightarrow$ Dark (D) $\rightarrow$ Dark + Whisker Trimming (DW) $\rightarrow$ Light + Whisker Trimming (LW).
  • Goal: To assess the impact of removing visual cues first, followed by tactile cues, and then the recovery potential of visual cues alone.
• Environment 2 (Tactile Cue-Enriched): An open-field arena in complete darkness containing salient tactile landmarks (sandpaper strips on walls and a square patch on the floor) but no visual cues.
  • Protocol: Mice established spatial maps in darkness. A cue-rotation experiment (90° rotation of tactile landmarks) was performed to test anchoring. Subsequently, whisker trimming was applied (DW) to test reliance on tactile input.
  • Advanced Manipulations: In a subset of mice, the sandpaper floor cue was removed (DWR), and finally, the arena walls were removed entirely (DWR'), leaving only the floor cue, to test the limits of boundary-dependent coding.
• Whisker-Movement Correlation: A separate cohort of mice underwent head-fixed sessions with "spontaneous" and "stimulated" whisking (using a sandpaper-tipped rod) while recording MEC activity to identify neurons directly correlated with whisker kinematics.
• Analysis:
  • Cell Classification: HD cells, border cells, grid cells, and spatial cells were identified based on standard metrics (Mean Vector Length, Border Score, Grid Score, Spatial Information).
  • Stability Metrics: Cross-session correlations, population vector (PV) correlations, and Linear Discriminant Analysis (LDA) were used to quantify representation stability.
  • Decoding: A Naïve Bayes decoder was used to assess position reconstruction accuracy based on population activity.

3. Key Contributions

• Differential Sensory Dependence: The study reveals that different MEC cell types have distinct hierarchies of sensory dependence. Border cells are uniquely sensitive to tactile input regardless of visual conditions, whereas grid and spatial cells are primarily visual-dependent in cue-rich environments but become tactile-dependent when vision is absent.
• Tactile Anchoring: It demonstrates that in the absence of vision, all classes of spatially modulated cells can anchor their firing fields to salient tactile landmarks, forming stable cognitive maps.
• Whisker-Specific Processing: The identification of a specific subset of MEC neurons whose activity correlates directly with whisker angular velocity, suggesting a direct neural pathway for integrating active tactile sensing into spatial navigation.
• Boundary Dominance: The study highlights that environmental boundaries (detected via whiskers or body contact) are the most critical stabilizing factor for spatial coding when all other sensory inputs are removed.

4. Key Results

A. Visual vs. Tactile Deprivation in Cue-Rich Environments (Env 1)

• HD Cells: Both light deprivation and whisker trimming significantly reduced tuning strength (MVL) and head information. The combination caused the most severe drift, and visual recovery (LW) did not fully restore the original population representation.
• Border Cells: Highly sensitive to both modalities. Whisker trimming caused a significant loss of border selectivity and increased firing in non-border regions, regardless of whether lights were on or off.
• Grid & Spatial Cells: Primarily dependent on vision. Light deprivation caused significant degradation of grid scores and spatial information. However, whisker trimming in the dark (DW) had minimal additional impact, suggesting that in a visually rich context, tactile cues are redundant for grid cell stability.

B. Tactile Anchoring in Darkness (Env 2)

• Cue Rotation: When tactile landmarks were rotated 90°, the firing fields of all cell types (HD, border, grid, spatial) rotated coherently by 90°, proving that tactile cues alone can anchor spatial maps.
• Whisker Trimming Impact: In this tactile-dependent environment, whisker trimming caused profound disruption across all cell types. Grid scores, spatial information, and decoding accuracy dropped significantly. This contrasts sharply with Env 1, confirming that grid cells rely on tactile cues when vision is unavailable.
• Field Loss: Fields located specifically over the tactile landmark (sandpaper) were more likely to disappear after whisker trimming than fields elsewhere.

C. The Critical Role of Boundaries

• When the sandpaper floor cue was removed (DWR), spatial coding remained relatively stable.
• However, when the walls were removed (DWR'), preventing physical contact with boundaries, spatial tuning collapsed across all cell types. Most cells lost their classification (e.g., grid cells lost hexagonal patterns). This indicates that boundary-derived tactile feedback (via nose/body contact) is the final, essential anchor for spatial maps when visual and active whisker cues are absent.

D. Whisker-Responsive Neurons

• Approximately 8.7% of recorded MEC neurons showed significant correlation with whisker angular velocity during stimulated whisking.
• Among these, a subset (33%) were also spatially modulated (mostly HD and border cells), while the majority were non-spatial.
• These neurons exhibited anatomical clustering, suggesting local circuitry for integrating tactile motion into spatial processing.

5. Significance

• Mechanism of Multisensory Integration: The paper provides a comprehensive map of how the MEC flexibly reweights sensory inputs. It challenges the notion that grid cells are purely path-integration based, showing they are highly dependent on external anchors (visual or tactile) to maintain regularity.
• Tactile Navigation: It establishes that rodents can construct robust spatial maps using tactile information alone, with whiskers playing a role analogous to vision in cue-rich environments.
• Neural Circuitry: The discovery of whisker-movement-correlated neurons in the MEC suggests a direct or indirect pathway (likely via somatosensory cortex $\rightarrow$ perirhinal/postrhinal cortices $\rightarrow$ MEC) that integrates active sensing with spatial cognition.
• Clinical/Behavioral Relevance: The findings explain how sensory deficits (e.g., blindness or loss of whiskers) might differentially impact navigation strategies and highlight the resilience of the brain's ability to switch sensory modalities, provided boundary cues remain accessible.

In summary, this study demonstrates that the MEC is a highly adaptable system that constructs spatial representations by integrating available sensory streams, with boundary contact serving as the ultimate stabilizer for spatial coding when other cues are compromised.