🧬 biology

Residual Depolarization as a Model of Associative Learning-Like Behavior in the Protozoan Stentor coereuleus

This paper proposes and validates a computational model based on residual membrane depolarization to explain associative learning-like behavior in the unicellular protozoan *Stentor coeruleus*, demonstrating that such complex behavior can arise from plausible non-synaptic cellular mechanisms.

Original authors: Hamza Mohamed

Published 2026-07-15✓ Author reviewed ⓘ

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Hamza Mohamed

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 ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed.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 a tiny, single-celled creature called Stentor coeruleus. It's about the size of a grain of sand, has no brain, no nerves, and no synapses. Yet, this little blob just pulled off a trick that scientists thought only animals with nervous systems could do: it learned to associate two different things.

Think of it like a magic trick. If you tap a creature gently (a "weak tap"), it usually ignores you. If you tap it hard (a "strong tap"), it jumps in fear. But here's the weird part: if you tap it gently, wait a second, and then tap it hard, the next time you tap it gently, it jumps even harder than before. It's as if the gentle tap became a warning signal for the big shock.

For a long time, scientists believed this kind of "associative learning" required a brain to connect the dots. But this paper suggests that Stentor might be doing it with something much simpler: a lingering electrical buzz in its cell membrane.

The "Ghost" in the Machine

The authors propose a model based on residual depolarization. Let's use an analogy to understand this.

Imagine the Stentor's cell membrane is like a trampoline.

The Resting State: Usually, the trampoline is flat. A gentle tap (the weak stimulus) isn't enough to make you bounce off the ground (contract).
The Trace: When you give that gentle tap, it leaves a tiny, invisible "ghost" of a bounce—a slight wobble that fades away quickly.
The Big Jump: If you hit the trampoline hard (the strong stimulus) while that ghost wobble is still there, the two forces combine. You get a massive bounce.
The Residual Effect: After that big bounce, the trampoline doesn't go back to being perfectly flat. It stays slightly bouncy for a while, like a trampoline that's been jumped on too many times. This is the "residual depolarization."

Because the trampoline is now slightly bouncy, the next time you give it a gentle tap, that tiny wobble is enough to make it jump! The creature has "learned" that the gentle tap means a big shock is coming, even though it has no brain to remember it. Crucially, the "jump" happens because the starting point (the baseline) has shifted closer to the edge, not because the edge itself moved. The paper emphasizes that the contraction threshold remains fixed; the cell just gets closer to it.

What This Paper Rules Out (The "Not-So-Simple" Explanations)

Before settling on this "bouncy trampoline" theory, the authors had to make sure they weren't being fooled by simpler tricks. They ran several control experiments to cross off other possibilities:

It's not just "getting excited": They tested if a single big shock just made the creature jittery and more likely to jump at anything. They found that if you gave a big shock and then a gentle tap later, the gentle tap didn't cause an extra jump. So, it's not just general arousal.
It's not "getting used to the shock": They wondered if repeated big shocks just made the creature more sensitive overall. But when they gave three big shocks in a row without the gentle taps, the gentle taps still didn't cause an extra jump.
It's not just "timing": They tried pairing two gentle taps together (weak-weak) to see if the timing alone caused the jump. It didn't. The special "inverted-U" shape of the learning (jumping up, then coming down) only happened with the weak-strong pairing.
It's not "boredom": They even "pre-habituated" the creatures by tapping them gently for a long time until they stopped reacting. Then, they tried the weak-strong trick again. Even though the creatures were bored with the gentle taps, the trick still worked!

These tests suggest that the behavior isn't a simple side effect; it's a specific reaction to the pairing of the two taps.

The Computer Simulation: A Virtual Test Tube

Since we can't stick electrodes into a single-celled organism easily enough to see every tiny electrical shift, the authors built a computational model. Think of this as a video game simulation of a Stentor.

They programmed the game with their "bouncy trampoline" rules:

A gentle tap leaves a short-lived trace.
A strong tap turns that trace into a lingering "bounciness" (residual depolarization).
Over time, the creature gets "tired" of being tapped (habituation), which eventually makes it stop jumping, even if the trampoline is still bouncy.

When they ran the simulation, the virtual Stentor behaved exactly like the real ones. It showed the same "inverted-U" curve: the response went up, peaked, and then went down.

Crucially, the paper suggests that this mechanism is necessary within the simulation to reproduce the observed data. When they "turned off" the residual depolarization in the simulation (by setting the "bounciness" update to zero), the learning effect disappeared. When they turned off the "tiredness" (habituation), the creature didn't stop jumping later on. This implies that in the model, both the lingering buzz and the eventual fatigue are required to get the results that match the real-world experiments.

How Sure Are We?

The authors are careful not to claim they have "solved" the mystery of the Stentor brain (which it doesn't have). Instead, the results suggest that a plausible cellular mechanism—residual depolarization—can reproduce the observed behavior.

The paper explicitly states that the model reproduced the trajectories from the experimental data and that the mechanisms were required in the simulation to get those results. However, these are simulations based on the data from Doan et al. (2026). The paper offers testable predictions for future experiments, such as measuring the actual membrane potential to see if it really does stay "bouncy" after the pairing.

The Takeaway

This paper doesn't say Stentor is thinking like a human. It suggests that a single cell might have a built-in, non-synaptic way to link events together. It's like a trampoline that remembers it was just jumped on, even without a brain to tell it so. The "learning" is just physics and chemistry playing out on a tiny, bouncy stage.

The authors found that the specific pairing of a weak tap followed by a strong tap creates a temporary state of high alertness that fades away, and their computer model proves this idea is physically possible. It's a fascinating glimpse into how complex behaviors might have started long before brains ever evolved.

Technical Summary: Residual Depolarization as a Model of Associative Learning-Like Behavior in Stentor coeruleus

Problem Statement
Associative learning is traditionally defined as a process requiring a nervous system and synaptic plasticity, where a conditioned stimulus (CS) paired with an unconditioned stimulus (US) elicits a preemptive response. However, recent observations by Doan et al. (2026) demonstrated that the unicellular protozoan Stentor coeruleus exhibits associative learning-like behavior: when weak mechanical stimuli (CS) are repeatedly paired with strong mechanical stimuli (US), the organism's response to the weak stimulus is transiently enhanced. This phenomenon challenges the assumption that associative learning necessitates a nervous system. While Stentor is known to undergo habituation (a decrease in response), standard habituation models cannot explain the observed increase in responsiveness above baseline. The central problem addressed by this study is identifying a plausible, non-synaptic intracellular mechanism that can reproduce the specific trajectory of this enhancement and subsequent decline, dependent on the specific pairing of weak and strong stimuli.

Methodology
The study employs a computational modeling approach to test the "residual membrane depolarization" hypothesis. The methodology proceeds through the following stages:

Data Preprocessing and Trajectory Analysis:
- Raw experimental data from Doan et al. (2026) regarding cell contractions (binary 0/1 responses) under Weak-Strong (WS) and Weak-Weak (WW) pairing conditions were analyzed.
- Population-level trajectories were quantified using Generalized Estimating Equations (GEE) logistic regression and aggregate binomial Generalized Linear Models (GLM). These statistical tests confirmed that WS and WW conditions exhibit significantly different trajectory shapes (specifically, an inverted-U trajectory for WS vs. a monotonic decline for WW), rather than just differing in average response probability.
Computational Model Construction:
A mechanistic model was developed in Python to simulate the cellular dynamics. The core hypothesis posits that a weak stimulus leaves a short-lived "trace" of subthreshold depolarization. If a strong stimulus follows while this trace persists, it converts the trace into a decaying, elevated baseline excitability state (residual depolarization).
- Contraction Probability: Modeled as a sigmoid function of the relative membrane potential ( $V_0 + \text{Baseline} + \text{Stimulus Strength} - \text{Threshold}$ ).
- Residual Depolarization ( $B(t)$ ): A state variable that updates only when a weak stimulus trace ( $T_w$ ) coincides with a strong stimulus response ( $P_s$ ). This update is subject to saturation limits and decays over the inter-trial interval (ITI).
- Habituation Mechanisms: The model incorporates two habituation pathways to explain the eventual decline in response:
  - Direct Weak-Pathway Habituation: Accumulation of load from repeated weak stimuli.
  - Cross-Habituation: Accumulation of load from strong stimuli affecting the weak pathway.
- Effective Stimulus Strength: Both weak and strong stimulus strengths are modeled as decreasing functions of their respective habituation loads.
Simulation and Fitting:
- The model was fitted to the first 15 trials of the WS and WW protocols (ISI = 1s, ITI = 45s) using bounded differential evolution to minimize the negative log-likelihood of the observed binomial contraction counts.
- Validation: The model's performance was evaluated through:
  - Trajectory Reproduction: Visual and statistical comparison of simulated curves against empirical data.
  - Ablation Analysis: Setting specific parameters (e.g., $B_{update}$ or $k_{cross}$ ) to zero to test the necessity of residual depolarization and cross-habituation mechanisms.
  - Control Protocols: Simulating unpaired sequences (Weak-only, Strong-Weak) and varying timing conditions (ISI/ITI variations) to ensure the behavior relies on specific pairing.
  - Pre-habituation Control: Simulating pre-habituated populations to rule out inherent salience of weak stimuli.
Electrophysiological Prediction:
The model generated testable predictions regarding the relative membrane potential dynamics during the WS protocol, visualizing the shift in baseline excitability before and after the strong stimulus.

Key Contributions and Results

Reproduction of Trajectories: The computational model successfully reproduced the distinct response trajectories observed in the experimental data. It captured the non-monotonic "early enhancement" (inverted-U shape) in the WS condition and the monotonic decline in the WW condition.
Mechanistic Necessity: Ablation analysis confirmed that the model's core mechanisms are required to reproduce the data. Specifically, removing the residual depolarization update ( $B_{update} = 0$ ) eliminated the early enhancement, while removing cross-habituation ( $k_{cross} = 0$ ) prevented the late decline, indicating that both mechanisms are essential for the full behavioral profile.
Specificity of Pairing: Control simulations demonstrated that the enhancement is strictly dependent on the specific temporal pairing of weak and strong stimuli. Unpaired sequences (Weak-only or Strong-Weak) failed to produce the enhancement, ruling out non-specific arousal or cumulative sensitization.
Robustness to Pre-habituation: The model accounted for the persistence of enhancement even in pre-habituated populations, supporting the hypothesis that the mechanism is distinct from simple responsiveness to salient stimuli.
Electrophysiological Predictions: The model provides a quantitative framework for the relative membrane potential, predicting a transient elevation of the baseline excitability state following the weak-strong pairing, which shifts the cell closer to the contraction threshold without altering the threshold itself.

Significance
The paper claims that the observed associative learning-like behavior in Stentor coeruleus can be fully reproduced by plausible, non-synaptic cellular mechanisms. By formalizing the "residual depolarization" hypothesis, the study suggests that a precursor to associative learning may exist in unicellular organisms, driven by the interplay of membrane potential dynamics and habituation processes rather than synaptic plasticity. The work provides a mechanistic explanation for how a single cell can exhibit complex temporal integration and learning-like behavior, offering a testable electrophysiological framework for future experimental verification.