🧠 neuroscience

Coordination of turn-related activity in the superior colliculus with locomotor dynamics and hippocampal representations of possible futures

This study reveals that turn-related activity in the superior colliculus is tightly coordinated with both the animal's stepping rhythm and hippocampal representations of future paths, enabling seamless navigation toward planned destinations.

Original authors: Wilhite, C., Frank, L. M., Scanziani, M.

Published 2026-08-12

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Wilhite, C., Frank, L. M., Scanziani, M.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine your brain as a bustling city where different neighborhoods have very specific jobs. Some neighborhoods are like the city's GPS, constantly mapping out where you are and where you could go next. Others are like the traffic control center, sending out the actual orders to your legs to take a step or turn a corner. For a long time, scientists thought these two neighborhoods worked in separate shifts: the GPS would figure out the plan, and then, once the plan was set, the traffic control center would just execute the moves. But what if they were actually having a constant, high-speed conversation while you were walking? What if the "GPS" was whispering suggestions to the "traffic cop" right in the middle of a step, and the "traffic cop" was timing its moves to the rhythm of your walking? This is the big question neuroscientists are asking: how does the brain's internal map of "what might happen next" talk to the muscles that actually move us?

This paper dives into that conversation by looking at two specific brain areas in mice: the hippocampus, which is famous for its "mental maps" of space, and the superior colliculus (SC), a midbrain structure that acts like a command center for turning your head and body left or right. The researchers were curious about two things. First, does the brain's turning signal sync up with the rhythm of the mouse's footsteps? Second, does the brain's "what-if" thinking (imagining a path before taking it) actually change how the turning command is fired? They wanted to see if the plan and the action are truly woven together in real-time, or if they happen in separate steps.

The Rhythm of the Turn

The researchers set up a Y-shaped maze for mice, a classic puzzle where the mouse has to choose to go left or right to find a tasty treat. As the mice scurried through, the team recorded the electrical activity of neurons in the superior colliculus, specifically looking for "turn cells"—neurons that fire when the mouse decides to turn left or right.

Here is the first big discovery: these turn cells are not just waiting for the mouse to decide to turn; they are dancing to the beat of the mouse's footsteps. The team found that the mouse's head bobs left and right in a rhythmic pattern as it walks, linked to its stepping cycle. They discovered that the "left-turn cells" in the brain fire at a specific moment in this rhythm, while the "right-turn cells" fire at the exact opposite moment. It's like a dance where the left-turners step on the beat when the right foot swings, and the right-turners step when the left foot swings.

This wasn't just happening at the moment of the turn. Even when the mouse was walking straight down a long hallway, these cells kept firing in their specific rhythm. This suggests that the brain is constantly preparing for a turn, keeping the left and right options ready in a rhythmic alternation, waiting for the perfect moment in the walking cycle to execute the move. It's as if the brain is saying, "I'm ready to turn left on the next beat, or right on the beat after that," ensuring the turn is smooth and stable.

The "What-If" Whisper

The second part of the story involves the hippocampus, the brain's GPS. We know that when animals approach a decision point, their hippocampus doesn't just show where they are; it sometimes "sweeps" ahead, simulating possible future paths. It's like a mental rehearsal where the mouse imagines running down the left arm of the maze, then the right arm, before actually choosing one.

The researchers found that these mental rehearsals directly influence the turn cells in the superior colliculus. When the hippocampus "swept" toward the left (imagining a left turn), the left-turn cells in the SC fired more vigorously. When the sweep went right, the right-turn cells fired more. This happened even before the mouse actually committed to the turn.

Crucially, the paper shows a one-way street in this conversation. The mental simulations in the hippocampus predict the activity in the turning center. However, the turning center's activity does not reliably predict the mental simulations. In other words, the "what-if" thinking seems to be the boss here, nudging the motor commands to favor the path being imagined. If the mouse is mentally rehearsing a left turn, the brain's turning command gets a boost for the left side, making the actual turn slightly tighter and more direct toward that imagined path.

The Takeaway

So, what does this all mean? The paper suggests that turning isn't just a mechanical reaction to a wall or a sudden decision. It is a highly coordinated event where the brain's internal map of possible futures talks directly to the motor commands that move the body. The "GPS" (hippocampus) whispers about possible paths, and the "traffic cop" (superior colliculus) listens, adjusting its rhythm to match the mouse's footsteps and amplifying the signal for the path being imagined.

The researchers found that about 30% of the neurons they looked at were these special "turn cells," and roughly half of those were locked into the stepping rhythm. They also saw that about a third of these cells changed their firing based on the direction of the mental sweep. The study doesn't claim to have solved the entire mystery of navigation, but it provides strong evidence that the brain's planning and its moving parts are deeply intertwined, working in a synchronized dance to help an animal navigate its world seamlessly. The mental image of a future path doesn't just sit in the background; it actively shapes the physical act of turning.

Technical Summary: Coordination of Turn-Related Activity in the Superior Colliculus with Locomotor Dynamics and Hippocampal Representations

Problem Statement
Animal navigation relies on the seamless execution of left and right turns, which must be precisely coordinated with the ongoing locomotor stepping rhythm and guided by internal plans for future paths. While the superior colliculus (SC), specifically its intermediate and deep motor layers (dSC), is known to contain neurons that report turn direction and trigger turning movements, it remains unknown how this turn-related activity is coordinated with two critical factors during freely moving navigation: (1) the rhythmic stepping cycle of locomotion, and (2) internal representations of possible future paths generated by the hippocampus. Previous studies have largely focused on movement-restricted animals or responses to external stimuli, leaving a gap in understanding how dSC activity unfolds during internally generated spatial navigation.

Methodology
The authors recorded neural activity from the dSC of freely moving mice navigating a Y-maze task designed to induce repeated left and right turns at a bifurcation. In a subset of experiments, simultaneous recordings were obtained from the dorsal CA1 region of the hippocampus.

Behavioral Tracking: Locomotor dynamics were tracked using high-speed video. For stepping analysis, a transparent linear track with a bottom-view camera allowed for the direct observation of forepaw swings. In the Y-maze, a top-view camera tracked head-body angle oscillations, which served as a proxy for the stepping rhythm (validated by the contralateral coupling of forepaw swings and head movements).
Neural Recording: Silicon probes were used to record from dSC "turn cells" (neurons showing significant direction selectivity) and hippocampal place cells.
Decoding and Analysis:
- Stepping Phase: Head-body angle oscillations were decomposed into phases (0 to $2\pi$ ) to determine the timing of turns relative to the stepping cycle.
- Hippocampal Sweeps: A state-space algorithm decoded the animal's position from hippocampal CA1 activity. "Sweeps" were identified as moments where the decoded position represented a future path (left or right arm) ahead of the animal's actual location.
- Statistical Modeling: Generalized Linear Models (GLMs) were employed to assess predictive relationships between hippocampal sweeps (occurring in theta cycles prior to the turn) and dSC firing rates at the bifurcation, and vice versa.

Key Contributions and Results

Identification of Turn Cells: Approximately 28% of recorded dSC neurons were identified as "turn cells," exhibiting significant firing rate modulation for specific turn directions (left or right). These cells predominantly fired for contraversive turns and became directionally selective roughly 300 ms before the animal reached the maze bifurcation, often before the behavioral trajectory diverged.
Coupling to Stepping Rhythm:
- The study established that mice execute left and right turns at opposite phases of the stepping cycle (left turns near phase 0/ $2\pi$ , right turns near phase $\pi$ ).
- Approximately 47% of turn cells exhibited firing significantly phase-locked to the stepping cycle.
- Crucially, left-preferring and right-preferring turn cells fired at opposite phases of the stepping cycle. This rhythmic alternation was maintained even during straight locomotion, suggesting that the preferred firing phase relative to the step cycle reveals a cell's turn preference even outside of the turn itself.
Coordination with Hippocampal Representations:
- Hippocampal sweeps representing future paths were detected as the animals approached the bifurcation. Most sweeps (ipsi-sweeps) aligned with the path the animal eventually took, while a minority (contra-sweeps) represented the alternative path.
- The activity of roughly one-third of turn cells was modulated by the direction of these hippocampal sweeps. Specifically, "sweep-selective" turn cells fired more when the hippocampus represented a future path matching the cell's preferred turn direction and less when it represented the non-preferred direction.
- This modulation was bidirectional: firing increased for preferred sweeps and decreased for non-preferred sweeps relative to trials without sweeps.
Temporal Predictive Relationship:
- Using GLMs, the authors demonstrated that hippocampal sweeps occurring up to four theta cycles (~500 ms) prior to the bifurcation could significantly predict dSC firing rates at the time of the turn.
- There was a positive correlation between the timing of these predictive sweeps and the time at which individual turn cells became directionally selective; earlier predictive sweeps corresponded to cells that became selective earlier.
- Conversely, dSC activity did not reliably predict the direction of subsequent hippocampal sweeps, suggesting a unidirectional flow of information from cognitive representations (hippocampus) to motor output (SC) during online navigation.
Behavioral Correlates: The direction of hippocampal sweeps predicted subtle deviations in the animal's behavioral trajectory. Trials with contra-sweeps (representing the unchosen path) resulted in tighter turns and greater deviations from a linear path compared to ipsi-sweep trials, suggesting that the consideration of alternative futures partially influences motor execution.

Significance
The paper claims that turn-related activity in the superior colliculus is highly integrated with both the organism's ongoing locomotor behavior and its internal cognitive plans. The findings suggest a mechanism where:

Motor Coordination: The SC aligns turn commands with specific phases of the stepping cycle, potentially creating windows of opportunity to execute turns efficiently and stably.
Cognitive-Motor Integration: The SC acts as an integration site where hippocampal representations of possible futures modulate motor commands. This supports the hypothesis that cognitive processes (mental simulations of paths) can "leak" into overt motor circuits to bias behavior, even when those paths are not ultimately selected.
Information Flow: The study proposes an asymmetry in information flow during online behavior, where cognitive representations (hippocampal sweeps) bias motor activity (SC), contrasting with potential reverse dynamics during offline states (e.g., sleep).

The authors conclude that this coordination ensures turns occur at optimal stepping phases and provides a neural pathway for cognitive circuits to guide behavior along internally planned trajectories.

The Rhythm of the Turn

The "What-If" Whisper

The Takeaway

Technical Summary: Coordination of Turn-Related Activity in the Superior Colliculus with Locomotor Dynamics and Hippocampal Representations

More like this