Introduction: Single-Cell Transcriptomics in Weakly Electric Fish

Questions

How does testosterone simultaneously change the electric organ, knollenorgans, and corollary discharge circuit?
Why can’t bulk RNA-seq tell us which cell types are responsible for gene expression changes in the electric organ?
What questions can single-nucleus RNA-seq answer in this system that previous approaches could not?

Objectives

Trace the mormyrid sensorimotor loop from command nucleus to electric organ to knollenorgan and back.
Summarize what testosterone does to each component of the loop and by what mechanism.
Distinguish the direct (corollary discharge circuit) from indirect (knollenorgan retuning) mechanisms of hormonal coordination.
Explain why single-nucleus RNA-seq is the natural next step after bulk transcriptomics in this system.

Lecture Video

Overview

In this module you’ll analyze real single-nucleus RNA-seq data from Brienomyrus brachyistius — a weakly electric fish from West Africa — to ask which cell types in the electric organ change gene expression when testosterone elongates the electric organ discharge (EOD). The lecture video above introduces the full biological context. This page provides the key concepts in brief and the exercises to work through after watching.

Electric fish are an unusually tractable system for linking gene expression to behavior. The EOD is produced by a dedicated peripheral motor system, detected by specialized electroreceptors, and gated by a precisely-timed corollary discharge circuit in the brain — and all three respond to the same hormonal signal. That makes it possible to trace a behavioral change (a longer EOD) to morphological changes, physiological changes, and molecular changes in a single, well-mapped system.

The gap single-nucleus RNA-seq fills is cell-type resolution. Prior work used bulk RNA-seq, which averages gene expression across all the cell types in the tissue. Single-nucleus RNA-seq decomposes that average, letting us ask which genes change in which cells.

The Sensorimotor Loop

When a mormyrid fires its electric organ, the command originates in the command nucleus (CN) in the hindbrain, travels via spinal electromotor neurons, and synchronously activates hundreds of flat, coin-shaped cells called electrocytes in the electric organ. The resulting EOD waveform depends directly on electrocyte membrane area and ion channel composition: more anterior-face membrane area raises capacitance and delays spike initiation, producing a longer waveform (Bass, Denizot & Marchaterre 1986; Bass & Volman 1987; Freedman et al. 1989).

The incoming signal from another fish is detected by knollenorgans (KOs) — electroreceptors specialized for conspecific communication signals. Raman & Hopkins (2025) showed these are precisely tuned to the species’ own EOD frequency spectrum, functioning as matched filters. When testosterone elongates the EOD, KO tuning shifts to match — but through an indirect mechanism: the changed EOD itself drives receptor cell turnover (Bass & Hopkins 1984).

The brain keeps the loop coherent using a corollary discharge (CD) pathway: a copy of each motor command routes from CN through the bulbar (BCA) and mesencephalic (MCA) command-associated nuclei to the sublemniscal nucleus (slem), which inhibits the electrosensory nucleus (nELL) just as the fish’s own EOD arrives there — making the fish effectively deaf to its own signal (Bell & Grant 1989). Fukutomi & Carlson (2023) showed this CD timing window shifts under testosterone to match the elongated EOD — and it does so directly, without requiring altered sensory feedback. Jarzyna & Carlson (2026) further showed that MCA is a conserved temporal hub: the same circuit node adjusts CD timing across seasonal (days), developmental (years), and evolutionary (millions of years) timescales.

Component	What testosterone does	Mechanism
Electric organ (electrocytes)	Elongates EOD waveform	Direct androgen target; membrane remodeling and ion channel retuning
Knollenorgans (KOs)	Shifts best frequency to match new EOD	Indirect; stimulus-driven receptor cell turnover
Corollary discharge (MCA)	Delays and elongates the inhibition window	Direct androgen target; independent of sensory feedback

Challenge 1: Stop and Predict (3 min)

Testosterone elongates a male’s EOD and shifts the spike that knollenorgan receptors fire in response to it. Before reading further: what would have to happen centrally in the brain to keep the fish from confusing its own (now-longer) EOD with an EOD from another fish?

The corollary discharge inhibition window must shift in both its onset and its duration to match the delayed and elongated reafferent input. Fukutomi & Carlson (2023) showed exactly this: CD inhibition timing tracks EOD duration precisely. Critically, the shift is a direct hormonal effect on the MCA circuit — not a consequence of experiencing a different EOD.

What’s Missing: Cell-Type Resolution

Bass & Hopkins (1983) established the seasonal dimorphism; the work above traced it across all three components of the loop. The next mechanistic layer is gene expression. Losilla & Gallant (2025) used bulk RNA-seq on B. brachyistius caudal peduncle tissue to identify 44 genes tracking EOD elongation — a coordinated shift in cytoskeletal, extracellular matrix, lipid-metabolism, and ion channel genes that aligns with the morphological and physiological data.

The limitation is resolution. Caudal peduncle tissue contains electrocytes, connective tissue, support cells, blood vessel cells, and immune cells. Bulk RNA-seq averages across all of them. A K⁺ channel gene that shifts in bulk data might be changing specifically in electrocytes — or in a support cell population with no role in action potential generation. Without cell-type resolution, mechanistic interpretation is limited. That is the gap single-nucleus RNA-seq fills, and it is the central motivation for this module.

Challenge 2: Putting It Together (10 min)

Sketch the full sensorimotor loop linking testosterone, the electric organ, knollenorgan electroreceptors, and the corollary discharge pathway. For each component, note (a) one change under testosterone treatment and (b) one open question that single-nucleus RNA-seq could address but bulk RNA-seq cannot.

Solution

Possible answers — the goal is to make the integration explicit, so yours may differ:

Electric organ / electrocytes — change: anterior-face membrane area expands; K⁺ and Na⁺ channel expression shifts toward longer APs. Open question: which of the 44 differentially expressed genes change specifically in electrocytes vs. connective tissue or immune cells?
Knollenorgan electroreceptors — change: best frequency shifts to match the elongated EOD. Open question: what transcriptional program drives KO receptor cell turnover, and is the same gene regulatory network active here as in the electric organ?
Corollary discharge (MCA) — change: onset and duration of CD inhibition delay and elongate. Open question: which MCA cell types express androgen receptors, and what genes change in those cells under testosterone?
Loop integration — change: all three systems stay internally coherent as EOD duration changes across seasons, development, and evolution. Open question: is there a shared androgen-responsive gene regulatory program operating across electrocytes, KO receptor cells, and MCA neurons?

Keypoints

The mormyrid sensorimotor loop links command nucleus → electric organ → knollenorgans → corollary discharge circuit, with a copy of each motor command used to predict and cancel self-generated sensory input.
Testosterone coordinately reshapes all three components — the electric organ (direct), knollenorgans (indirect, stimulus-driven), and the corollary discharge circuit (direct, MCA-centered) — keeping the loop coherent as the EOD changes.
The MCA is a conserved temporal hub: the same circuit node adjusts CD timing across seasonal, developmental, and evolutionary timescales.
Bulk RNA-seq on electric organ tissue identified 44 candidate genes but cannot resolve which genes changed in which cell types.
Single-nucleus RNA-seq is the natural next step: it applies to frozen archived tissue and provides cell-type-resolved gene expression across the full sensorimotor circuit.

--- title: "Introduction: Single-Cell Transcriptomics in Weakly Electric Fish" teaching: 5 exercises: 15 optional: false --- ::: {.callout-note title="Questions"} - How does testosterone simultaneously change the electric organ, knollenorgans, and corollary discharge circuit? - Why can't bulk RNA-seq tell us which cell types are responsible for gene expression changes in the electric organ? - What questions can single-nucleus RNA-seq answer in this system that previous approaches could not? ::: ::: {.callout-tip title="Objectives"} - Trace the mormyrid sensorimotor loop from command nucleus to electric organ to knollenorgan and back. - Summarize what testosterone does to each component of the loop and by what mechanism. - Distinguish the direct (corollary discharge circuit) from indirect (knollenorgan retuning) mechanisms of hormonal coordination. - Explain why single-nucleus RNA-seq is the natural next step after bulk transcriptomics in this system. ::: ## Lecture Video ```{=html} <video class="plyr-video" playsinline controls preload="metadata"> <source src="https://d18e7eu8nurr5a.cloudfront.net/lectures/introduction_v2.mp4" type="video/mp4" /> Your browser does not support embedded video. </video> ``` ## Overview In this module you'll analyze real single-nucleus RNA-seq data from *Brienomyrus brachyistius* — a weakly electric fish from West Africa — to ask which cell types in the electric organ change gene expression when testosterone elongates the electric organ discharge (EOD). The lecture video above introduces the full biological context. This page provides the key concepts in brief and the exercises to work through after watching. Electric fish are an unusually tractable system for linking gene expression to behavior. The EOD is produced by a dedicated peripheral motor system, detected by specialized electroreceptors, and gated by a precisely-timed corollary discharge circuit in the brain — and all three respond to the same hormonal signal. That makes it possible to trace a behavioral change (a longer EOD) to morphological changes, physiological changes, and molecular changes in a single, well-mapped system. The gap single-nucleus RNA-seq fills is cell-type resolution. Prior work used bulk RNA-seq, which averages gene expression across all the cell types in the tissue. Single-nucleus RNA-seq decomposes that average, letting us ask which genes change in which cells. ## The Sensorimotor Loop When a mormyrid fires its electric organ, the command originates in the **command nucleus (CN)** in the hindbrain, travels via spinal electromotor neurons, and synchronously activates hundreds of flat, coin-shaped cells called **electrocytes** in the electric organ. The resulting EOD waveform depends directly on electrocyte membrane area and ion channel composition: more anterior-face membrane area raises capacitance and delays spike initiation, producing a longer waveform ([Bass, Denizot & Marchaterre 1986](files/introduction/Bass_etal_1986_androgen_binding_brain_electric_organ.pdf); [Bass & Volman 1987](files/introduction/Bass_etal_1987_behavior_to_membranes_testosterone_AP_duration.pdf); [Freedman et al. 1989](files/introduction/Freedman_etal_1989_temporal_analysis_testosterone_electric_organs.pdf)). The incoming signal from another fish is detected by **knollenorgans (KOs)** — electroreceptors specialized for conspecific communication signals. [Raman & Hopkins (2025)](files/introduction/Raman_Hopkins_2025_high_frequency_tuning_knollenorgan.pdf) showed these are precisely tuned to the species' own EOD frequency spectrum, functioning as matched filters. When testosterone elongates the EOD, KO tuning shifts to match — but through an **indirect mechanism**: the changed EOD itself drives receptor cell turnover ([Bass & Hopkins 1984](files/introduction/Bass_Hopkins_1984_androgen_shifts_knollenorgan_tuning.pdf)). The brain keeps the loop coherent using a **corollary discharge (CD)** pathway: a copy of each motor command routes from CN through the bulbar (BCA) and mesencephalic (MCA) command-associated nuclei to the sublemniscal nucleus (slem), which inhibits the electrosensory nucleus (nELL) just as the fish's own EOD arrives there — making the fish effectively deaf to its own signal ([Bell & Grant 1989](https://doi.org/10.1523/JNEUROSCI.09-03-01029.1989)). [Fukutomi & Carlson (2023)](files/introduction/Fukutomi_etal_2023_hormonal_coordination_motor_sensory_electric_fish.pdf) showed this CD timing window shifts under testosterone to match the elongated EOD — and it does so **directly**, without requiring altered sensory feedback. [Jarzyna & Carlson (2026)](files/introduction/Jarzyna_etal_2026_shared_neural_substrates_sensorimotor_integration.pdf) further showed that MCA is a conserved temporal hub: the same circuit node adjusts CD timing across seasonal (days), developmental (years), and evolutionary (millions of years) timescales. | Component | What testosterone does | Mechanism | |---|---|---| | Electric organ (electrocytes) | Elongates EOD waveform | Direct androgen target; membrane remodeling and ion channel retuning | | Knollenorgans (KOs) | Shifts best frequency to match new EOD | Indirect; stimulus-driven receptor cell turnover | | Corollary discharge (MCA) | Delays and elongates the inhibition window | Direct androgen target; independent of sensory feedback | ::: {.callout-important title="Challenge 1: Stop and Predict (3 min)"} Testosterone elongates a male's EOD and shifts the spike that knollenorgan receptors fire in response to it. Before reading further: what would have to happen *centrally* in the brain to keep the fish from confusing its own (now-longer) EOD with an EOD from another fish? :::: {.callout-solution title="Solution" collapse="true"} The corollary discharge inhibition window must shift in both its onset and its duration to match the delayed and elongated reafferent input. Fukutomi & Carlson (2023) showed exactly this: CD inhibition timing tracks EOD duration precisely. Critically, the shift is a direct hormonal effect on the MCA circuit — not a consequence of experiencing a different EOD. :::: ::: ## What's Missing: Cell-Type Resolution [Bass & Hopkins (1983)](https://doi.org/10.1126/science.6844924) established the seasonal dimorphism; the work above traced it across all three components of the loop. The next mechanistic layer is gene expression. [Losilla & Gallant (2025)](files/introduction/Losilla_etal_2025_gene_expression_EOD_duration_mormyrid.pdf) used bulk RNA-seq on *B. brachyistius* caudal peduncle tissue to identify 44 genes tracking EOD elongation — a coordinated shift in cytoskeletal, extracellular matrix, lipid-metabolism, and ion channel genes that aligns with the morphological and physiological data. The limitation is resolution. Caudal peduncle tissue contains electrocytes, connective tissue, support cells, blood vessel cells, and immune cells. Bulk RNA-seq averages across all of them. A K⁺ channel gene that shifts in bulk data might be changing specifically in electrocytes — or in a support cell population with no role in action potential generation. Without cell-type resolution, mechanistic interpretation is limited. That is the gap single-nucleus RNA-seq fills, and it is the central motivation for this module. ::: {.callout-important title="Challenge 2: Putting It Together (10 min)"} Sketch the full sensorimotor loop linking testosterone, the electric organ, knollenorgan electroreceptors, and the corollary discharge pathway. For each component, note (a) one change under testosterone treatment and (b) one open question that single-nucleus RNA-seq could address but bulk RNA-seq cannot. :::: {.callout-tip title="Solution" collapse="true"} Possible answers — the goal is to make the integration explicit, so yours may differ: - **Electric organ / electrocytes** — change: anterior-face membrane area expands; K⁺ and Na⁺ channel expression shifts toward longer APs. Open question: which of the 44 differentially expressed genes change specifically in electrocytes vs. connective tissue or immune cells? - **Knollenorgan electroreceptors** — change: best frequency shifts to match the elongated EOD. Open question: what transcriptional program drives KO receptor cell turnover, and is the same gene regulatory network active here as in the electric organ? - **Corollary discharge (MCA)** — change: onset and duration of CD inhibition delay and elongate. Open question: which MCA cell types express androgen receptors, and what genes change in those cells under testosterone? - **Loop integration** — change: all three systems stay internally coherent as EOD duration changes across seasons, development, and evolution. Open question: is there a shared androgen-responsive gene regulatory program operating across electrocytes, KO receptor cells, and MCA neurons? :::: ::: ## Keypoints 1. The mormyrid sensorimotor loop links command nucleus → electric organ → knollenorgans → corollary discharge circuit, with a copy of each motor command used to predict and cancel self-generated sensory input. 2. Testosterone coordinately reshapes all three components — the electric organ (direct), knollenorgans (indirect, stimulus-driven), and the corollary discharge circuit (direct, MCA-centered) — keeping the loop coherent as the EOD changes. 3. The MCA is a conserved temporal hub: the same circuit node adjusts CD timing across seasonal, developmental, and evolutionary timescales. 4. Bulk RNA-seq on electric organ tissue identified 44 candidate genes but cannot resolve which genes changed in which cell types. 5. Single-nucleus RNA-seq is the natural next step: it applies to frozen archived tissue and provides cell-type-resolved gene expression across the full sensorimotor circuit.