A galanin-positive population of lumbar spinal cord neurons modulates sexual arousal and copulatory behavior in male mice

Lenschow, Constanze; Mendes, Ana Rita P.; Ferreira, Liliana; Lacoste, Bertrand; Marques, Hugo; Gutierrez-Castellanos, Nicolas; Quilgars, Camille; Bertrand, Sandrine S.; Lima, Susana Q.

doi:10.1038/s41467-025-63877-2

Download PDF

Article
Open access
Published: 23 September 2025

A galanin-positive population of lumbar spinal cord neurons modulates sexual arousal and copulatory behavior in male mice

Nature Communications volume 16, Article number: 8282 (2025) Cite this article

7912 Accesses
1 Citations
207 Altmetric
Metrics details

Subjects

Abstract

During sex, male arousal builds to the ejaculatory threshold, allowing genital sensory input to trigger ejaculation. While copulation and arousal are thought to be brain-regulated, ejaculation is a reflex controlled by a spinal circuit. In this framework, the spinal cord is assumed to be strongly inhibited by descending input until the ejaculatory threshold, playing no role in the regulation of copulatory behavior. However, this remains untested. Here we mapped the spinal circuit controlling the bulbospongiosus muscle, essential for sperm expulsion in mice. Our findings show that bulbospongiosus muscle-motor neurons receive input from galanin-expressing neurons, which integrate genital sensory signals. Stimulating these neurons induces bulbospongiosus activity, but responses vary with spinalization, internal state, and decrease with repeated stimulation. Ablating galanin-positive neurons altered ejaculation latency and copulatory patterns. These results suggest that spinal circuits influence not only ejaculation but also copulation and arousal, challenging the traditional view of spinal control in copulation.

Low frequency neuromuscular electrical stimulation applied to the bulbospongiosus muscle prolongs the ejaculation latency in a rat model

Article 13 February 2023

A transcriptome dataset for gonadectomy-induced changes in rat spinal cord

Article Open access 29 December 2022

Rare phenomena of central rhythm and pattern generation in a case of complete spinal cord injury

Article Open access 06 June 2023

Introduction

Sexual behavior typically begins with pre-copulatory actions, such as courtship displays, that increase sexual motivation (e.g., erection in males and receptivity in females). In species with internal fertilization copulatory actions, including direct genital stimulation, bring the male to the ejaculatory threshold: at this point penile insertion in the female’s reproductive organ triggers ejaculation and sperm ejection¹. This is usually followed by a variable period of decreased male sexual activity^2,3,4.

While pre-copulatory and copulatory actions are primarily regulated by the brain, in coordination with the autonomic and sensory-somatic nervous systems^5,6,7, ejaculation in contrast is a two-step reflex controlled by autonomic and somatic circuits within the spinal cord. During the first step, called emission, sperm and seminal fluids are released and accumulate in the prostatic urethra^8,9. The second step, expulsion, is a somatic reflex caused by the activation of spinal motor neurons (MNs) and the subsequent contraction of the bulbospongiosus muscle (BSM), a large, striated muscle surrounding the base of the penis^10,11,12,13.

How copulatory sequences are orchestrated and how the arrival to the ejaculatory threshold is communicated to the spinal cord allowing genital input to trigger the reflex, remains unresolved. Key insights come from experiments in anesthetized male rats, where stimulation of the penis (via electrical stimulation of the dorsal penile nerve) was shown to elicit ejaculation in spinalized animals, but not intact ones^14,15. While this result strongly supports the idea that genital stimulation is sufficient to trigger the ejaculatory reflex, it also suggests that descending inhibitory signals to the spinal cord may modulate the impact of incoming sensory input, preventing inadvertent ejaculations^16,17. This contributes to the view of the brain as both a spinal-reflex inhibitor, and a central organizer of sexual behavior^15,18,19,20. In this model, arrival at the ejaculatory threshold may be communicated to the spinal circuitry by the transient interruption of descending inhibition, allowing genital stimulation to activate the ejaculatory reflex⁵. Ablation of the rat spinal interneurons controlling the BMS-MNs (a key element of the circuitry controlling the ejaculatory reflex in rats), fully prevented ejaculation, while leaving other aspects of copulatory behavior intact²¹, suggesting a minimal role of the spinal ejaculatory network in the organization of pre-copulatory and copulatory behaviors. However, this model remains incomplete, partly because most studies were performed either in humans, where manipulations are limited, or in rats, using methods with poor anatomical and cellular specificity, and/or low temporal resolution.

Here, we took advantage of the house mouse, a species whose sexual behavior consists of copulatory patterns that resemble human sexual dynamics, in particular with respect to repeated vaginal thrusting preceding ejaculation (in contrast to rat sexual behavior, where ejaculation is dependent on the execution of multiple individual penile insertions, spaced in time, each one with the potential of triggering an ejaculation^4,22,23). To gain access to the circuitry controlling the ejaculatory reflex, we focused on the pelvic muscle responsible for sperm expulsion, the BSM, as a point of entry into the spinal cord. Employing genetic-based strategies with high specificity (viruses combined with transgenic mice) and high temporal resolution (optogenetics and electrophysiology) we identified the mouse BSM-MNs and their presynaptic partners, a group of galanin-positive (Gal+) interneurons in the lumbar spinal cord. We provide functional evidence that the Gal+ neurons receive genital sensory input and that their stimulation evokes dominant BSM activity, but only after spinalization. Surprisingly, the evoked BSM activity depends on the sexual excitement of the male prior to spinalization, indicating that the spinal cord circuitry controlling BSM activity reflects the animal’s internal state. While Gal+ neurons appear to be active prior to ejaculation, genetic ablation of this population significantly impacts the length and structure of copulatory behavior. These results point towards an unexpected and more intricate role of the spinal cord during copulation, extending beyond the simple relay of genital information and the control of the ejaculatory reflex.

Results

Characterization of the bulbospongiosus muscle motor neurons (BSM-MNs)

To visualize the MNs involved in sperm expulsion, we targeted the BSM to identify the corresponding BSM-MNs in the spinal cord. Fluorogold (FG), a well-established retrograde tracer²⁴ was injected into the BSM of adult mice (N = 12 mice, Fig. 1A, B). Serial rostrocaudal reconstruction of the spinal cords revealed a consistent distribution of FG-positive (FG+) BSM-MNs in the dorsomedial part of the ventral horn, near the dorsal gray commissure (Fig. 1B). FG+ somas were observed across several spinal segments, spanning from the lumbar 3 (L3) to the sacral 2 (S2) segments (Supplementary Fig. 1), with the majority of cells located between lumbar 6 (L6) and sacral 1 (S1) segments (Fig. 1C). Immunohistochemical characterization of the FG+ somas revealed that BSM-MNs are classic alpha MNs, expressing choline acetyltransferase (ChAT) and osteopontin, which are alpha MN markers²⁵. These neurons also receive sensory afferent input, as indicated by the presence of Vesicular Glutamate Transporter 1 (VGLUT1) positive boutons (Fig. 1D). Nearly all FG+ BSM-MNs expressed ChAT (75 out of 76 FG+ cells quantified) and VGLUT1 (76 out of 76 FG+ cells) and 80% were osteopontin positive (61 out of 76 FG+ cells quantified; Fig. 1E). The soma size of FG+ MNs was similar to that classically reported for alpha MNs (Fig. 1F)²⁶.

**Fig. 1: Anatomical and functional characterization of the bulbospongiosus motor neurons (BSM-MNs).**

To establish a causal relationship between the activity of BSM-MNs and muscle activity, we employed optogenetic methods to selectively activate these MNs²⁷ while simultaneously monitoring BSM activity with electromyography (EMG; Fig. 1G–M). We injected a retrograde traveling adeno-associated virus (rAAV), carrying the gene for the light activated channel channelrhodopsin (ChR2; rAAV-CAG-ChR2tdTomato, Addgene²⁸), into the BSM of postnatal day 3–6 (P3-P6) mouse pups (Fig. 1G, upper panel), as the efficiency of viral particles to infect motor endplates is known to drastically drop at later postnatal days²⁹, and animals were raised until sexual maturity (2–3 months of age). To activate the ChR2-expressing somas, a laminectomy was performed to allow blue light illumination of several BSM-MNs-containing spinal segments. To assess the specificity of the light stimulation, activity was recorded from both the BSM and a locomotor hindlimb muscle, the Tibialis Anterior (TA) (N = 10 mice, Fig. 1G). BSM EMG potentials were tightly locked to laser stimulation (Fig. 1H BSM trace, Supplementary Fig. 2). The largest amplitudes (Fig. 1I), shorter latencies (Fig. 1K), and lowest light intensities (Fig. 1J), were obtained with illumination above the L6 and S1 segments, in agreement with the BSM-MN rostrocaudal density (Supplementary Fig. 2). No activity was observed in the TA (Fig. 1H, TA trace, Supplementary Fig. 2).

In five animals, we performed single-cell juxtacellular recordings from photoidentified BSM-MNs (N = 7 cells; Fig. 1L, MN trace³⁰) alongside BSM EMG recordings. Brief blue light stimulation (10 ms, 15 mW) at different frequencies (upper panel: 5 Hz, middle panel: 10 Hz, lower panel: 20 Hz) reliably elicited short latency action potentials (mean latency to spike: 4.643 ms +/− 0.33 ms, Fig. 1M), followed by BSM EMG potentials (Fig. 1H). Spike and EMG fidelities (calculated as the number of spikes or EMG responses divided by the number of light pulses) remained stable up to 20 Hz (Fig. 1N). Moreover, light evoked BSM-MN activity resulted in characteristic pelvic floor movements that were tightly locked to the laser (Supplementary Movie 1), resembling the movements observed during ejaculation in a sexually behaving male (Supplementary Movie 2)²². Together, these results revealed a population of MNs, primarily located in the L6 and S1 spinal segments, whose optogenetic activation induces characteristic EMG potentials in the BSM and ejaculatory-like movements of the pelvic floor.

Characterization of the BSM-MNs presynaptic partners

We next aimed to identify the presynaptic partners of BSM-MNs. To do so, we started by injecting the BSM of adult mice with pseudorabies virus (PRV, Kaplan strain^31,32,33) (Fig. 2A). Similar to the FG and retro-AVV injections, PRV-labeled neurons were primarily found in the dorsomedial part of the ventral horn of the L6 segment (Fig. 2B). In the rat, BSM-MNs have been reported to receive input from galanin-expressing (Gal+) interneurons located around the central canal in the lamina X of the L3/L4 spinal segments²¹. Consistent with these findings, dense PRV labeling was observed around the central canal in the L2/L3 spinal segments that overlapped with post-hoc immunohistochemical labeling of galanin (Fig. 2C), revealing the existence of a similar population of Gal+ cells in the mouse.

**Fig. 2: The BSM-MNs receive direct input from a group of lumbar spinal cord Galanin positive (Gal+) neurons.**

To specifically access the Gal+ population, we utilized a mouse line where Cre recombinase is expressed under the control of the galanin promoter (Gal-cre mice)^34,35. Gal-cre mice were crossed with a reporter mouse line carrying the gene for the red fluorescent protein tdTomato³⁶, resulting in progeny expressing the fluorescent protein in Gal+ neurons (Gal-tdT+ , Fig. 2D). Immunohistochemical staining for galanin in the progeny from the Gal-cre x tdTomato cross confirmed the specificity of this mouse line (Fig. 2D & Supplementary Fig. 3). TdTomato-positive neurons were observed along the rostral caudal axis of the spinal cord, namely in laminae X (around the central canal, N = 10 mice, Fig. 2E), with the highest cell density located in the L2/L3 spinal segments (Fig. 2E). The mouse Gal+ neurons expressed enkephalin, cholecystokinin, gastrin releasing peptide and substance P, similar to the rat Gal+ neurons^37,38,39, as all four peptides were present and overlapped with the Gal+ cells surrounding the central canal in the L2/L3 spinal segments (Supplementary Fig. 4).

To determine if the results obtained with the PRV injections are due to the existence of a monosynaptic connection between the Gal+ cells and the BSM-MNs³³, we took a two-pronged approach. First, we aimed to identify the location of the synaptic terminals of the lumbar Gal+ cells by injecting a Cre-dependent AAV carrying a green fluorescent protein (GFP)-tagged form of synaptophysin (a synaptic vesicle protein, present in neuronal terminals⁴⁰) into the L2/L3 spinal segments of Gal-cre x tdTomato mice (Fig. 2F, N = 7 mice). To visualize the BSM-MNs, mice were simultaneously injected with FG in the BSM (Fig. 2F). GFP-labeled terminals belonging to the Gal+ cells (green channel, Fig. 2G) were found around FG+ BSM-MNs (blue channel, Fig. 2G) in all 7 animals (Fig. 2H). On average, a total of 79% ± 2% FG+ labeled cells overlapped with synaptophysin-labeled terminals (Fig. 2H) indicating that the Gal+ neurons in the L2/L3 spinal segments contact the BSM-MNs directly. Furthermore, this approach revealed other projection targets of the Gal+ cells, including the intermediolateral column (IML) and the central autonomic nucleus (CAN) in the lower thoracic and L1 spinal segments (Supplementary Fig. 5), as well as the sacral parasympathetic nucleus (SPN) (Supplementary Fig. 5). The IML and CAN are known to contain sympathetic preganglionic cells that provide sympathetic outflow mainly through the hypogastric and pelvic nerves to the visceral organs^41,42,43,44. The parasympathetic preganglionic cells clustered in the SPN at the S2-S5 spinal segments (Supplementary Fig. 5) innervate the pelvic organs, including the prostate⁴⁵, urethra^46,47 and penis⁴⁷. The dorsolateral nucleus (DLN) located at the S1/S2 segments, known to consist of MNs innervating the ischiocavernosus muscle, the most important muscle for erection⁴⁸, also contained GFP-labeled terminals of the Gal+ cells. All the regions containing GFP-labeled terminals were also labeled after PRV injection in the BSM (Supplementary Fig. 6), except for the DLN which contains ischiocavernosus MNs. These results suggest that the areas sending information to the Gal+ cells receive reciprocal input, as previously described in the rat⁴⁹, and confirm that the PRV initial infection was restricted to the BSM.

To obtain physiological evidence supporting a direct synaptic connection between the Gal+ cells and the BSM-MNs, we performed in vitro whole-cell patch clamp recordings from cholera toxin-B (CTB-594) retrogradely-labeled BSM-MNs in acute spinal cord slices of mouse pups (P2-P6) expressing ChR2 under the galanin promoter (Gal-Chr2, progeny from the cross between the Gal-cre mouse and the ChR2 mouse)⁵⁰. It is important to note that using adult sexually mature animals for this approach is very challenging, as MNs undergo rapid cell death when performing spinal cord slices beyond a certain age, likely due to their complex dendritic branching⁵¹.

An optical fiber (0.2 mm diameter) was placed close to the central canal (Fig. 2I) to illuminate the terminals of the Gal+ cells. A total of 21 CTB-positive BSM-MNs were recorded in Gal-ChR2 pups (N = 7 mice), of which 18 reliably responded to laser illumination (Fig. 2J–M; Supplementary Fig. 7A, B). Short laser pulses led to dominant excitatory postsynaptic potentials (EPSPs) and/or action potentials in CTB-positive BSM-MNs (see upper traces in Fig. 2L, M) also upon consecutive laser application (Supplementary Fig. 7D–F). Repeating laser stimulation after the superfusion of a high cation containing aCSF, to isolate monosynaptic transmission⁵², resulted in much smaller EPSPs in BSM-MNs (middle traces in Fig. 2L, M; Supplementary Fig. 7C). Blocking neural transmission via the pharmacological manipulation of NMDA (through AP5) and AMPA (through DNQX) receptors (lower traces in Fig. 2L, M) abolished light-triggered EPSPs in BSM-MNs. In contrast to the BSM-MNs, other large CTB-negative lumbar MNs (N = 16 cells) did not respond to laser illumination (1 out of 16 responded, Fig. 2K, N). Taken together, these findings provide evidence for a population of Gal+ neurons present in the laminae X of the L2/L3 spinal segments in mice, which is monosynaptically connected to BSM-MNs via glutamatergic transmission.

Sensory input to the Gal+ neurons and the BSM-MNs

It has been hypothesized that once the ejaculatory threshold is reached, genital input can activate the spinal circuitry controlling the ejaculatory reflex⁵. This model is supported by anatomical evidence, as previous studies have reported connections between the sensory branch of the pudendal nerve and putative rat Gal+ neurons^53,54. However, to the best of our knowledge, functional connectivity has never been established. Therefore, next we investigated whether the BSM-MNs and Gal+ cells receive sensory feedback from the penis.

We first replicated an experiment previously conducted in rats^54,55 (where the penis is stimulated electrically or through air puffs) in male mice that were either intact (Supplementary Fig. 8A) or spinalized (Supplementary Fig. 8C) while monitoring the BSM activity in parallel through EMG. Consistent with findings in rats, we observed prominent BSM EMG activity when stimulating (200 Hz, 100 pulses, 6 V; 5 Hz, 3 × 5 pulses of 100 ms, 6 V) the penis in a spinalized preparation, whereas penile stimulation in intact mice resulted in minimal BSM activation (Supplementary Fig. 8B–D). Penile stimulation-triggered BSM responses were significantly larger in amplitude (Supplementary Fig. 8E) and duration (Supplementary Fig. 8F) in spinalized mice compared to non-spinalized mice, although the onset of BSM responses did not differ (Supplementary Fig. 8G). These experiments further support the existence of supraspinal inhibition of the spinal circuitry, but do not yet confirm whether penile sensory information is physiologically integrated at the level of Gal+ cells and/or BSM-MNs.

To address this question, we used an optogenetic approach and mapped light-induced local field potentials (LFP) of ChR2 infected BSM-MNs in the lumbar spinal cord of adult male mice combined with BSM EMG recordings (Fig. 3A). Prominent time-locked LFP deflections were observed when illuminating above the BSM-MNs-containing spinal segments (Fig. 3B, blue trace, Vm) which were followed by BSM activity (Fig. 3B, black trace, BSM EMG). After mapping the optimal location of laser induced BSM-MN LFP deflections and BSM EMG activity, the position of the glass pipette capturing LFP deflections was maintained while applying brief air puffs of 5 Hz (10 ms, 1 bar) to the pulled-out penis, or the leg as a control (Fig. 3C). Air puff stimulations led to larger LFP deflections when applied to the penis (Fig. 3C, left panel, blue trace) compared to leg puffs (Fig. 3C, middle panel, yellow trace), a result that was consistent at the population level (Fig. 3C, right lower panel). Plotting the light-induced BSM-MNs LFPs against the LFP deflections induced by penis puffs (blue) or the leg puffs (yellow) showed a higher correlation between the light-induced LFPs and the penis puffs (Fig. 3C, right upper panel). These results hint to the BSM-MNs receiving sensory input from the penis.

**Fig. 3: Gal+ neurons and BSM-MNs receive sensory input from the penis.**

We repeated the same experiment in male mice expressing ChR2 under the Gal promoter (Gal-ChR2), as we mapped light-induced LFP signals at the location of the Gal+ cells (Fig. 3D) while monitoring BSM activity in a spinalized preparation. Optogenetic stimulation at the L2/L3 spinal segments evoked prominent BSM responses (Fig. 3E, black trace) in line with time-locked LFP deflections (Fig. 3E, pink trace). We then performed air puff stimulations on the penis or leg while maintaining the LFP capturing pipette at the spot of the most effective light-induced LFPs (Fig. 3D, right panel). Penile sensory stimulation elicited significantly larger LFP responses (Fig. 3F, left panel, pink trace) compared to leg air puffs (Fig. 3F, middle panel, yellow trace and right lower panel). Moreover, light-induced LFPs were more correlated with LFPs induced by penis puffs than those elicited by leg puffs (Fig. 3F, right upper panel) indicating that penile sensory inputs can reach the Gal+ cells present at the L2/L3 spinal segments. These findings provide evidence that both the BSM-MNs and the Gal+ population are the recipients of penile sensory information.

Description of electrically induced BSM activity

To further explore the role of the Gal+ neurons in the control of the BSM activity in mice, we investigated whether the artificial activation at their location could trigger BSM activity. First, we performed electrical stimulations along the rostrocaudal lumbar spinal cord by inserting a tungsten electrode at various depths (550−850 μm), while concurrently measuring BSM activity using EMG recordings (Fig. 4A) in sexually naive, spinalized animals (SN, N = 8 mice). We simultaneously monitored the activity of the TA leg muscle to assess the specificity of our protocol (Fig. 4A).

**Fig. 4: Electrical stimulation of the lumbar spinal cord location harboring the Gal+ neurons leads to BSM activity in an anesthetized preparation.**

The site with the strongest electrically triggered BSM responses was labeled by electrolytic lesions and immunohistochemical staining for galanin (Fig. 4B, upper panel). Successful BSM muscle activity was elicited (Fig. 4B and Supplementary Fig. 9) when stimulations (40 μA, 200 Hz, 100 pulses) were applied near the Gal+ cluster, located in the L2/L3 spinal segments (Fig. 4C), at a depth of 850 μm, which aligns with the location of the central canal (Fig. 4D and Supplementary Fig. 9C,D). The electrically triggered BSM responses were significantly higher and longer in spinalized preparations compared to non-spinalized ones (Supplementary Fig. 10), suggesting descending inhibition from the brain. This finding aligns with our previous results from the mechanical/electrical stimulation of the penis experiments (Fig. 3) and prior studies in rats^16,17.

However, unlike what was observed in the rat, where repeated electrical stimulations led to consistent BSM responses⁵⁶, repeated stimulations in mice resulted in decreased BSM responses (see “Methods”; Fig. 4B; BSM 2nd and 3rd). Quantification of the EMG responses (Fig. 4E, F) showed significantly larger and longer potentials during the first train of stimulation (mean amplitude 0.99 mV +/− 3.6 mV, 1st vs 2nd amplitude p = 0.04, 1st vs. 3rd amplitude p = 0.02; mean length 8.83 ms +/− 1.4 ms; 1st vs 2nd length p = 0.09, 1st vs. 3rd length p = 0.02; Mann-Whitney-U Test) compared to the second (mean amplitude 0.5 mV +/− 0.25 mV; mean length 4.22 s +/− 1.47 s) and third trains (mean amplitude 0.06 mV +/− 0.04 mV; mean length 1.74 s +/− 1.31 s) of current application, while the onset of the responses did not change (mean onset of 1st 3.9 +/− 1.83 s, mean onset of 2nd 6.03 +/− 3.02 s, mean onset 3rd 0.54 +/− 0.13 s; Fig. 4G). The reduction in response during repeated stimulation was not due to deterioration of the preparation (Supplementary Fig. 11).

In addition to the intriguing depressing response observed with repeated electrical stimulations, the structure of the electrically elicited BSM activity in mice differed from previously recorded EMG responses during ejaculation in anaesthetized rats (see ref. ⁵⁶). Moreover, we did not observe the expulsion of sperm or seminal fluid in any of our mouse experiments. To rule out the hypothesis that the lack of sperm emission in mice was due to a technical issue, we replicated the experiment by Borgdorff et al. (2008)⁵⁶ in anesthetized rats. As expected, the application of electrical current (Supplementary Fig. 12A) at the L3/L4 spinal segments (as confirmed by electrolytic lesions; Supplementary Fig. 12B) resulted in the characteristic activity pattern in the rat BSM (N = 5 rats; Supplementary Fig. 12C) and the expulsion of sperm (Supplementary Fig. 12D).

To explore how electrically induced BSM activity relates to natural physiological conditions, we conducted in vivo EMG recordings of BSM activity in sexually behaving mice (Supplementary Fig. 13A). BSM activity was detected at various stages during the sexual interaction (Supplementary Fig. 13B). Notably, distinct BSM contractions were observed with each male thrust (Supplementary Fig. 13C), with the strongest bursts occurring as the penis exited the vagina. In contrast, no BSM contractions were present during probing behavior (the period after mounting and before penile intromission, when the male is performing shallow thrusts, trying to locate the female’s vagina, Supplementary Fig. 13D). These findings support the hypothesis that the BSM might be involved in maintaining an erection (as the BSM is active during intravaginal thrusts) but not in initiating it.

Interestingly, no BSM activity was detected during the shuddering phase, the brief period immediately following the last thrust leading to ejaculation, characterized by rapid pelvic movements and lasting less than 3 seconds⁵⁷ (Supplementary Fig. 13B). Since ejaculation consists of two stages—emission and expulsion—we hypothesize that the shuddering phase corresponds to the emission phase, during which sperm is deposited in the urethra while the penis is less erect⁵⁸. The subsequent expulsion of sperm is driven by BSM contractions observed immediately after the shuddering phase (Supplementary Fig. 13B, E). This hypothesis, suggesting that sperm emission occurs during shuddering and expulsion afterward, is supported by previous studies in mice⁵⁷, which found that separating males from females during this phase prevented pregnancy and led to sperm expulsion at the tip of the penis outside the vagina.

As expected, the BSM activity recorded immediately after the shuddering phase showed the highest EMG activity, with bursts averaging a mean amplitude of 0.732 mV, lasting more than 6 seconds (Supplementary Fig. 13F). Based on the same separation studies in mice⁵⁷ sperm expulsion is thought to occur during the first 1–2 s immediately after the shuddering phase occurs. Therefore, after sperm expulsion, the post-shuddering phase activity is probably related to flipping and cupping movements of the penis⁵⁷. Importantly, the dynamics of BSM EMG activity in freely moving animals mirrored those elicited by direct optogenetic stimulation of the BSM-MNs (Fig. 1H) and by sensory stimulation of the penis (Supplementary Fig. 8D). Similar patterns were also observed during electrical stimulation of the location containing Gal+ neurons, including comparable amplitudes and durations, and oscillatory patterns (Supplementary Fig. 13G). These consistent results across experiments further support that we are observing expulsion-like patterns in the BSM, but no sperm emission, when artificially activating the Gal+ cells in anesthetized in vivo preparations.

As previously noted, mice and rats display distinct reproductive strategies, particularly in the number of ejaculations each species can achieve in a short period of time: while rats can ejaculate 7–8 times in a short period before reaching sexual exhaustion, mice, similar to humans, generally enter a refractory period after a single ejaculation, which in the case of C57BL6 mice (the genetic background used in this study) can last several days⁴. We speculated that the inability to elicit stable BSM activity with repeated electrical stimulations in mice might be due to the first round of stimulation inducing a refractory state similar to that following ejaculation.

To test this hypothesis, we allowed male mice to have sex prior to the electrical stimulation experiments. One group of males was allowed to reach ejaculation (Ejaculation, N = 8 mice; Fig. 4H upper panel), while another group was allowed to perform 5 mounts with intromission and intravaginal thrusting (5 Mounts, N = 7 mice, Fig. 4I upper panel). Electrical stimulation of the spinal cord in the Ejaculation group evoked less BSM activity (Fig. 4H, lower panel) compared to the Sexually Naive group (Fig. 4B, first train of stimulation), and the 5 Mounts group (Fig. 4I, lower panel). The mean amplitude of the BSM activity was significantly higher in the 5 Mounts group (Fig. 4J), while the duration of the BSM activity was significantly shorter in the Ejaculation group compared to the other two groups (Fig. 4K). Although there was no difference in the mean amplitude of the BSM events between the Sexually Naive and Ejaculation groups (Fig. 4J), the number of events was significantly lower in the Ejaculation group compared to the other two (Fig. 4M), further suggesting that after ejaculation, electrical stimulation of the spinal cord cannot elicit high BSM activity.

Taken together, these experiments suggest that electrical stimulation of the mouse spinal cord in the area where the Gal+ neurons are located can only evoke the second and final phase of the ejaculatory process (expulsion), but not the emission phase. The differing outcomes of the electrical stimulations based on the male’s internal state (Sexually Naive vs. 5 Mounts vs. Ejaculation) imply that the properties of neuronal circuits in this spinal cord region are modulated by copulation and ejaculation, suggesting that the spinal circuitry may be involved in controlling the refractory period in addition to ejaculation.

Optogenetic stimulation of Gal+ neurons leads to BSM activity

To address the inherent lack of specificity in electric stimulation experiments and to unequivocally link the activity of Gal+ cells, BSM-MNs and the BSM, we genetically restricted the neuronal population being stimulated to the Gal+ neurons by crossing the Gal-cre mouse line with the ChR2 line (Gal-ChR2⁵⁰). To activate the Gal+ neurons, we placed an optical fiber on top of the spinal cord of anesthetized spinalized mice and delivered brief pulses of blue light along its rostrocaudal axis (Fig. 5A, left panel, N = 10 mice). Light delivery (20 mW, 200 Hz, 100 pulses with 10 ms duration) led to comparable responses to the ones elicited via electrical stimulation (Fig. 5B lower right panel). As expected, BSM activity was only observed in response to illumination at the location with the highest ChR2 expression (revealed by either DiI injections or an electrolytic lesion, Fig. 5B, see also Supplementary Fig. 14), specifically above the L2/L3 spinal segments (Fig. 5C, D), the region previously shown to harbor the highest density of Gal+ cells (Fig. 2E). Consistent with the results obtained from electric stimulations, optogenetically-evoked BSM responses were markedly higher in amplitude and longer in duration in spinalized compared to intact preparations (Supplementary Fig. 15), further supporting the idea that the Gal+ cells may be subject to tonic descending inhibition^14,16,17,18.

**Fig. 5: Optogenetic activation of the Gal+ neurons leads to BSM activity in an anesthetized preparation.**

Furthermore, optogenetically elicited BSM responses exhibited similar dynamics to those evoked by electrical stimulation. Both methods resulted in a decrease in amplitude and duration with consecutive trains of stimulation (see example traces in Fig. 5B and population data in Fig. 5E, F), while the onset of triggered responses remained unchanged (Fig. 5G).

In addition to these bursts of activity, we used lower stimulation frequencies to trigger time locked BSM responses (N = 12 mice; 20 mW, 5–50 Hz, 15 pulses; Fig. 5H–M). In a subset of animals, we successfully conducted juxtacellular recordings in photo-identified single Gal-ChR2 neurons (Fig. H–M, N = 7 cells), revealing response dynamics upon optogenetic stimulation in individual Gal+ cells and the consequent dynamics of the BSM with time locked triggered activity. The mean latency to spike for Gal+ cells was 10.6 ms ± 1.7 ms, followed by light triggered BSM EMG with a mean latency of 6.7 ms ± 0.09 ms. Notably, the mean latency to spike for photo-identified BSM-MNs was 4.5 ms ± 0.4 ms; Fig. 5I), further supporting the existence of a monosynaptic connectivity between Gal+ cells and BSM-MNs, as demonstrated above through in vitro recordings (Fig. 2I–N). Spike and EMG fidelities (calculated as the number of spikes or EMG responses divided by the number of light pulses) were stable up to 10 Hz (Fig. 5J). Importantly and similar to the tonic discharges observed above, a decrease in amplitude (Fig. 5K) and duration (Fig. 5L) was observed with consecutive lower frequency optogenetic stimulation, with a constant onset of triggered responses (Fig. 5M).

Finally, the decrease in BSM activity observed with the electric and optogenetic stimulation of Gal+ cells contrasts sharply with the responses evoked by direct optogenetic activation of BSM-MNs. Direct optogenetic stimulation of BSM-MNs resulted in consistent responses in single BSM-MNs and in the BSM across repeated rounds of stimulation (Figs. 1G–N and 5N–P). This suggests that the depression in BSM activity upon consecutive stimulation of Gal+ neurons may be due to an alteration in synaptic transmission between the Gal+ cells and BSM-MNs, rather than changes in the intrinsic properties of BSM-MNs.

Genetic ablation of Gal+ neurons disrupts copulatory behavior

Given the unexpected finding that the outcome of electrical stimulation of the spinal cord on BSM activity was dependent on the behavioral state of the male, suggesting the involvement of Gal+ cells during other phases of copulatory behavior (see Fig. 4H–M), we used the expression of the immediate early gene cFos to establish a link between behavior and neuronal activity⁵⁹. For this purpose, we compared the induction of the cFos protein in the Gal+ cells of the L2/L3 spinal segments of male mice that ejaculated (sexual interaction with a receptive female until ejaculation, Ejaculation, N = 6 mice) with the induction in males that were either alone in a clean cage (Cage control, N = 7 mice) or had varying degrees of interaction with a female (10 min with a receptive female but no penile insertion as intromission was interrupted whenever the male attempted copulation, Sexually Excited, N = 5 mice; or five mounts with intromission and intravaginal thrusting, 5 Mounts, N = 5 mice) (Fig. 6A–D).

**Fig. 6: The lumbar population of Gal+ neurons becomes increasingly active during sexual behavior.**

To specifically quantify the number of active Gal+ cells, we used male progeny from the Gal-cre x tdTomato cross and performed post-hoc immunohistochemical quantification of the tdTomato signal and cFos induction in the L2/L3 spinal segments for each condition. As expected, given the direct contact to the BSM-MNs, significantly more double-positive neurons were found in the Ejaculation group, compared to the other three conditions (Cage control, p = 0.0000015, Sexually Aroused, p = 0.00011 and 5 Mounts, p = 0.002 Two-tailed Student’s t-test, Fig. 6E). Unexpectedly, we observed similar levels of activation between the Sexually Excited and 5 Mounts groups, with both conditions exhibiting significantly higher levels of double-positive neurons compared to the Cage control (p = 0.0067, Two-tailed Student’s t-test). To test if the activation observed in the Sexual Excited group was specific to an interaction with a female, we conducted a separate experiment evaluating cFos activation levels after introducing a male intruder in the cage with the test animal and compared it to the Cage control and Sexually Excited groups. The interaction with the intruder male, which involved bouts of aggressive behavior and male-directed mounting, elicited similar levels of cFos activation in the Male and Sexually Excited groups (Supplementary Fig. 16). Given the high arousal state of the male in these two conditions, this result suggests that the activity of the Gal+ cells may be related to the internal state of the male and not specific for a sexual interaction.

Our results so far suggest that the population of lumbar Gal+ cells might have a role before the ejaculatory threshold. To determine the function of this spinal population during a full sexual interaction, we used a genetic approach to specifically ablate the Gal+ cells by expressing the diphtheria toxin receptor (DTR⁶⁰). A conditional AAV carrying the DTR construct (AA8V-FLEX-DTR-GFP, Salk) was injected into the L2/L3 spinal segments of sexually trained males derived from the Gal-Cre x tdTomato cross (Fig. 7A; DTR group; N = 12 mice). The control group, of the same genotype and also sexually trained, underwent a sham surgery at the same spinal location (SHAM group; N = 7 mice).

**Fig. 7: Genetic ablation of the lumbar Gal+ neurons disrupts male sexual behavior.**

After a recovery period and a second session of sexual behavior, both groups (DTR and SHAM) received an intraperitoneal injection of diphtheria toxin (DT). One week after DT treatment, the effect of the genetic ablation on sexual behavior was tested in the presence of a sexually receptive female. Males of both groups were sacrificed either 90 min after ejaculation or 90 min after the female was introduced into the testing arena if the male initiated copulation but did not ejaculate. In cases with absent sexual motivation (no mount attempts), the trial was interrupted 30 min after the female was introduced, and testing was repeated up to two more times, once a week (Supplementary Fig. 17A and see “Methods” for detailed experimental procedure⁴).

Post-hoc immunohistochemical processing of the spinal cords (Fig. 7B, C) allowed us to determine the number of cFos-positive and Gal+ cells in the L2/L3 spinal segments. While the number of cFos-positive cells was comparable between the SHAM and DTR group (Fig. 7B–D), there was a significant reduction in the number of Gal+ cells (Fig. 7E, G) and in the overlap of Gal+ neurons and cFos-positive cells (Fig. 7F, G). Consistent with the previous cFos and electrical stimulation experiments (Figs. 3 and 5) and further supporting the involvement of these neurons in a more general control of sexual behavior, the copulatory sequence (Fig. 7H) was significantly disrupted in the DTR group. While 3 out of 12 DTR animals did not reach ejaculation (but attempted copulation), and only one ejaculated in less than 10 min, all SHAM animals reached ejaculation, with only two taking longer than 10 min (Fig. 7I), as indicated in the cumulative distribution of the latency to ejaculate from the first mount (Mount with probing—MP or Mount with Intromission—MI). Individual raster plots for each animal aligned to the first consummatory act (Fig. 7J) illustrate no effect on the latency to mount (Fig. 7K), while the latency to ejaculate from first mount was significantly longer (Fig. 7L).

To identify which aspect of the sexual interaction was disrupted, leading to an increased latency to ejaculate, we further analyzed the copulatory sequence (including only DTR animals that ejaculated in the analysis). Sexual motivation did not differ across groups as reflected by the latency to mount (Fig. 7K) and the number of anogenital investigations, though there was a trend towards a higher number of anogenital investigations in the DTR group (Supplementary Fig. 17B), in particular when considering the number of events during the consummatory phase of the behavior (after the first mount, Supplementary Fig. 17B).

Interestingly the mounting events resulting in penile insertion (MI, Fig. 7H), had similar durations across groups (Fig. 7M and Supplementary Fig. 17C), and the number of MI was also similar (Fig. 7N), indicating that a comparable amount of genital sensory information was sufficient to trigger ejaculation in both types of males, despite differences in the latency to ejaculate. In contrast, we observed a difference in the number of mounts that did not result in penile insertion: mount attempts where the male performed several shallow pelvic thrust movements trying to locate the female’s vagina (which only occur if the male is sexually motivated, MP; Fig. 7O⁶¹). Surprisingly, the ejaculation duration was similar across all males that ejaculated (from the last thrust until the male dismounted the female, see “Methods”; Fig. 7P), indicating that if the male reached the ejaculatory threshold, sperm expulsion could occur. This contrasts with results from similar experiments in rats, where ablation of rat Gal+ cells disrupted ejaculation, but not copulation²¹. In line with the increased number of MP, the time to successful penile insertion after the male placed his paws on the female flanks was also significantly longer in the DTR group (Fig. 7Q, Supplementary Fig. 17D). Thrusting rate and dynamics of MI were similar across groups (Fig. 7R–T), indicating normal pelvic thrusting once penile insertion was achieved. Importantly, erection did not seem affected by the ablation of Gal+ cells, as penile movements were similar in SHAM and DTR animals, with both groups capable of penile insertion and intravaginal thrusts (as observed in the Supplementary Movies S3 and S4).

Finally, to test if the behavioral impairments were due to the extent of the ablation, we assessed the relationship between the number of double-positive neurons (Gal+ and cFos+) and various behavioral parameters. There was no correlation between the number of MI, total number of thrusts, mount duration (Supplementary Fig. 18A, B), or other behavioral parameters (Supplementary Fig. 18C–J) and the double-positive neurons, meaning that the magnitude of behavioral impairments is not only explained by the extent of ablation, but perhaps by other compensatory mechanisms discussed below. Taken together, our results indicate that the ablation of the lumbar Gal+ population leads to an increase in the latency to ejaculate and a significant disruption of the copulatory sequence, supporting a more complex involvement of the spinal cord in controlling sexual behavior in mice.

Discussion

In mammals, pre-copulatory and copulatory actions orchestrated by the brain are thought to bring the male to the ejaculatory threshold, such that penile insertion in the female’s vagina triggers ejaculation and sperm ejection, a reflex controlled by a spinal cord circuitry. This assumption implies that the brain plays no role in the ejaculatory reflex (other than inhibiting it until the threshold is reached) and that the activity of the spinal network is inconsequential to the orchestration of copulatory behavior (serving only as a relay of ascending sensory input). However, this division of labor between the brain and the spinal cord has been repeatedly challenged, particularly by the existence of neurological disorders suggesting that spinal neurons are integral to the control system, with central and effector players in continuous information exchange^62,63. Our study provides evidence supporting the involvement of the spinal cord in controlling male mice’s internal state and copulation, rather than merely relaying penile stimulation and controlling ejaculation. Using the penile muscle involved in sperm expulsion as a point of entry, we have characterized a microcircuit in the lumbar spinal cord of male mice consisting of motor neurons innervating the bulbospongiosus muscle (BSM-MNs) and a population of galanin-positive (Gal+) neurons that is monosynaptically connected to the BSM-MNs. Moreover, we provide strong anatomical evidence for a reciprocal connection between the Gal+ neurons and autonomic nuclei, as well as the ischiocavernosus MNs. This suggests a potential role of the Gal+ population not only in emission but also in erection^45,64,65,66. The output to the ischiocavernosus MNs might be important for the regulation of penile detumescence, which must be finely regulated at the time of ejaculation, as pressure has to momentarily decrease for expulsion to take place⁶⁷.

The population described in our study shares a similar molecular profile with the previously identified rat Spinal Ejaculator Generator (SEG^37,38,39), despite some differences in the spinal segments where the Gal+ neurons are located (L2/L3 segments versus L3/L4 in the rat²¹). Several lines of evidence supported labeling the rat Gal+ population as the SEG. While electrical stimulation at the location of the putative SEG evokes ejaculation in anesthetized rats⁵⁶, its ablation results in complete disruption of the ejaculatory reflex without affecting copulatory behavior²¹. Furthermore, the SEG seems to be anatomically connected to the sensory branch of the pudendal nerve, indicating it received genital information^53,54. However, to the best of our knowledge, the functional connectivity between the rat SEG and the BSM-MNs has never been established.

Our study, using a dual and genetic-based approach combined with in vitro whole-cell patch-clamp recordings of BSM-MNs while specifically activating Gal+ axons, provides data unequivocally establishing a functional monosynaptic connection between Gal+ cells and BSM-MNs. Despite the molecular similarities between the Gal+ neurons described in this study and the rat SEG, our results indicate that the properties of the mouse spinal network may differ significantly from its rat counterpart.

First, electrically induced ejaculation in anesthetized rats produces similar motor and physiological activity patterns regardless of an intact connection to the brain (which we reproduced in our study), while penile stimulation can activate ejaculation only in anesthetized spinalized rats¹⁵. These findings suggest the presence of descending inhibition that gates sensory input from the penis and that the rat SEG can function independently of the brain once this inhibition is removed. In contrast, in mice, pronounced BSM activity was only triggered when the Gal+ cells were electrically or optogenetically activated in a spinalized preparation, but not if the connection to the brain was intact. Moreover, penile stimulation did not trigger an ejaculation in our experiments. Therefore, our results support a model where, at least in mice, the Gal+ population and the incoming sensory input is kept under a substantial inhibition from the brain, likely from the ipsilateral paragigantocellular thalamic nucleus²⁰, until the ejaculatory threshold is reached.

We were intrigued by our inability to trigger emission in mice through Gal+ and penile stimulation, despite demonstrating an anatomical connection to autonomic centers, the main players in the emission process⁶⁸, in contrast to what was observed in the rat. While we cannot rule out the possibility that our failure to trigger emission was due to suboptimal stimulation parameters, we propose that these results are biologically plausible and may reflect species differences in the regulation of ejaculation.

Even though we could not induce emission through artificial stimulation of the spinal cord location containing the Gal+ cells or by directly activating the Gal+ neurons, our EMG recordings in behaving animals showed that artificially-induced BSM activities closely resembled natural BSM activity, further supporting that our stimulation parameters elicited activity within physiological ranges and that Gal+ stimulation alone is insufficient to induce emission.

Mouse copulation is characterized by repeated vaginal thrusting leading to a single ejaculation, a copulatory sequence more similar to humans but quite different from rats, which have multiple ejaculations dependent on the execution of multiple individual penile insertions. This suggests that the buildup to the ejaculatory threshold in rats may differ from mice and humans, as it might be independent or much less influenced by the integration of genital sensory input, while in the mouse, sensory integration of multiple thrusts is crucial. While rat data suggest that descending inhibition operates primarily at the level of the SEG, our data points for the descending inhibition to modulate the Gal+ neurons and the incoming sensory input to them in mice.

While anatomically the circuitry seems to be similar between rats, mice and humans, the functional logic may be fundamentally different. The rat spinal circuitry appears to function as a true spinal reflex arc, capable of driving ejaculation with a sensory stimulus if disconnected from the brain. In contrast, the mouse circuit does not qualify as a proper reflex since the sensory input alone cannot drive emission or full BSM activity, supporting a constant dialogue between the brain and the spinal cord, for the buildup of sexual excitation and ejaculation. Future comparative experimental efforts are needed to reveal the operational logic of these circuits in different species. For example, it remains to be determined if a true central pattern generator for ejaculation exists in the mouse and how the Gal+ neurons we have identified connect to such network. However, the discrepancies outlined in this study may simply result from different modus operandi, the product of different evolutionary trajectories. We observed an unexpected gradual increase in the activity of the Gal+ population during sex and when interacting with another male, suggesting the activity of these neurons is related to the arousal level of the individual. This contrasts with previous studies in rats, which showed that Gal+ cells are only active after an ejaculation⁶⁹. To our surprise, repeated electrical stimulation was accompanied by a marked decrease in BSM activity, which was not due to a deterioration of the preparation. Moreover, if male mice were allowed to have sex and ejaculate just before the electric stimulation experiment, BSM activity was significantly lower and resembled the activity observed after multiple rounds of stimulation. These results indicate the Gal+ neurons are recruited not only during ejaculation but also in response to arousal and sexual activity. This supports the notion of a continuous dialogue between the spinal cord and the rest of the body and a possible prominent role of the spinal cord in the control of the refractory period³, contrary to what is currently hypothesized.

Lastly, the chemogenetic ablation of the Gal+ population caused a disruption in sexual behavior. Interestingly, and in contrast to the rat SEG, whose ablation led to a complete abolishment of ejaculation while leaving copulatory patterns intact²¹, only 3 out of 12 animals in our study did not ejaculate. However, since the session was artificially interrupted after a certain time limit, it is unknown if they would have eventually ejaculated given more time. Also, and contrary to the rat, we observed altered sexual behavior in all animals, including increased latency to ejaculate, more mounts with probing, and longer durations of probing, suggesting that immediate sensory penile feedback is partially disrupted when Gal+ cells are ablated.

These results raise two immediate questions: how can animals still ejaculate if the Gal+ neurons that contact the BSM-MNs are ablated, and how and where is the penile sensory feedback processed to reach ejaculation? The first question may be explained by a small percentage of Gal+ cells not being ablated, which could suffice to integrate pelvic sensory input and drive ejaculation. Additionally, redundant spinal circuits may compensate for the loss of Gal+ cells. Namely, circuits controlling pelvic floor muscles and organs, responsible for sexual behavior, micturition and defecation, share common connections between spinal and brain nuclei⁷⁰. For instance, connections between the sacral parasympathetic nuclei that control pelvic organ function (i.e., emission), and the BSM-MN nucleus have been described in cats^71,72 and rats⁴¹. Holstege and Tan (1987) hypothesized that BSM-MNs are a special class of neurons with both somatic and autonomic functions, capable of controlling both processes⁷².

Moreover, direct brain spinal projections, namely to the BSM-MNs, may drive sperm expulsion independently of Gal+ cells. Anatomical evidence indicates a direct connection between the hypothalamic paraventricular nucleus (PVN) and the BSM-MNs in cats⁷² and rats^73,74. This pathway appears to regulate erection but also rat sexual behavior in general⁷⁵. Additionally, in rats, ejaculation can be facilitated by an oxytonergic PVN projection onto SEG cells⁷⁶. Projections from the Barrington’s nucleus, also known as the pontine micturition center, connect to pelvic-related nuclei in the spinal cord (in rats⁷⁷, cats⁷⁸ and humans⁷⁹), and electrical stimulation of the medial-lateral areas induced emission and expulsion-like responses, respectively⁷⁸.

Regarding the sensory relay, studies in rats and humans showed that pressure increase in the prostatic urethra caused by the arrival of sperm and seminal fluid at the time of emission is sufficient to induce ejaculatory-like contractions in the BSM^80,81. Rat dorsal penile nerve to BSM-MNs connectivity has been described^82,83,84 and might be sufficient to relay sensory information, without the Gal+ neurons in our experiments. Finally, Qi et al. (2023) described Krause corpuscles as relayers of genital sensory information in mice, and their manipulation impacted sexual behavior, reinforcing the idea that sensory feedback may be necessary to sustain copulation and ejaculation⁸⁵. As for the second question, it remains unclear where arousal is controlled and how sensory input impinges on such circuits. But our results point towards an involvement of the spinal network. In summary, we have identified a cluster of Gal+ cells in the lumbar spinal cord of mice that are directly connected to BSM-MNs and that seem to be involved in the integration of sensory signals during copulation and the male’s internal state, thereby taking a more central and intricate involvement in the control of mice copulatory behavior than a simple reflex arc.

Methods

Experimental model and subject details

All experimental procedures were carried out in strict accordance with the guidelines of the European Committee Council Directive and were approved by the Animal Care and Users Committee of the Champalimaud Neuroscience Program, the Portuguese National Authority for Animal Health (Direcção Geral de Veterinária; approval number 0421/000/000/2022) and by the local ethic committee of the University of Bordeaux and the French Agriculture and Forestry Ministry for handling animals (approval number 2016012716035720). For tracing from the BSM (FG and PRV), electrical stimulation experiments and EMG recordings in behaving animals, C57BL6 (Mus musculus domesticus, C57BL/6J) male mice aged 3–6 months were used. For tracing from the spinal cord, the cFos experiment and the DTR ablation, Gal-cre x tdTomato (Mus musculus domesticus, B6;129S6-Gt(ROSA)26Sortm9(CAG-tdTomato)Hze/J) male mice aged 3–6 months old were obtained from Jackson Laboratories. For the spinal cord optogenetic stimulation and airpuff/electrical sensory stimulation of the penis, Gal- ChR2 (Mus musculus domesticus, B6;129S-Gt(ROSA)26Sortm32(CAG-COP4*H134R/EYFP)Hze/J) male mice aged 3–6 months old were obtained from Jackson Laboratories. For electric stimulation of the spinal cord, Long Evans (Crl:LE Long-Evans) male rats were ordered from Charles Rivers. For the pup muscle injections, C57BL6 male mice pups aged P3-P6 were used. For the pup optogenetic and electrophysiology experiment, Gal-ChR2 male mice pups aged P2-P6 were used. Finally, as a sexual stimulus, after ovariectomy and hormonal priming, C57BL6 female mice aged 3–6 months were used. All the animals were bred and maintained in our animal facility. Except for the optogenetic experiment in pups, all animals were weaned at 21 days and housed in same-sex groups in stand-alone cages (1284L, Techniplast, 365 × 207 × 140 mm) with access to food and water ad libitum. All experimental mice were maintained in suites with controlled temperature and humidity (22 celsius, 50% humidity), and on an inverse 12:12 light/dark cycle. Experiments were performed during the dark phase of the cycle, phase of higher animal activity. After initiation of sexual behavior training or immediately after surgery, male mice were kept single-housed until the experiment was over. All animals were randomly allocated into the different experimental groups tested in this study.

Bulbospongiosus muscle (BSM) injections

Mice were anesthetized with 3% isoflurane in oxygen and put into the mouthpiece of a stereotaxic device (Kopf, Tujunga, CA, USA). After that, mice were turned into a supine position to facilitate access to the BSM. After shaving the anogenital area and cleaning with Betadine (MEDAPharma) and 70% ethanol, a small incision in the scrotum area was made. At this point, an analgesic (buprenorphine, 0.05–0.1 mg/kg, intraperitoneal injection) was administered. The BSM was exposed after removing fat and conjunctive tissue. Pulled capillaries (length 3 1/2 inches [9 cm]; inner diameter 0.53, outer diameter 1.14 mm; tip diameter 40 μm; DrummondScientific, Broomall, PA, USA) were used to inject 1.5 μl of 2% Fluorogold (FG) at a rate of 13.8 nL per pulse, and a frequency of 0.2 Hz. In total, 5 BSM sites were injected with FG (three in the dorsal and two in the ventral portion of the BSM). A waiting time of 5 min prior to and 10 min after injection was kept. The glass pipette was pulled out slowly and the skin sutured. All mice were single housed post-surgery, for 1 to 2 weeks, until perfusion. A total of 12 male mice was used for anatomical investigation of BSM motor neurons using FG.

Another 6 male mice were injected with PRV-Ka-gEI-mCherry (PRV³²) in the BSM. The male was similarly anesthetized with 3% isoflurane in oxygen and put into the mouthpiece of a stereotaxic device. After that, mice were turned into a supine position to access the BSM, and after disinfecting, a small incision was made to expose the muscle. Using a glass pipette as described above, 1 μl of PRV was injected once into the BSM at a rate of 13.8 nL per pulse, and a frequency of 0.2 Hz. A waiting time of 5 min prior to and 10 min after injection was kept, after which the pipette was pulled and the incision sutured. All mice were single housed post-surgery, for 3–4 days, until perfusion.

Spinal cord stereotaxic viral injections and histology

Mice were anesthetized with 3% isoflurane in oxygen and spinally fixed into a stereotaxic frame using adapted vertebrae clamps (Kopf, Tujunga, CA, USA). During surgery, anesthesia was maintained using 1.5% isoflurane. Injection sites (lumbar segments 2/3) were targeted by using vertebral landmarks as described in ref. ⁸⁶. Muscle and fat tissue were gently removed to get a better sight onto the spinal cord. The injection pipette was inserted in between the thoracic vertebrae T11 and T12 before having made a small puncture into the dura mater that allowed for better insertion of the injection pipette. Pulled capillaries (length 31/2 inches [9 cm]; inner diameter 0.53, outer diameter 1.14 mm; tip diameter 40 μm; DrummondScientific, Broomall, PA, USA) were used to inject the AAVs at 100 μm medial from the midline, at a depth of 750–850 μm from the spinal cord surface at a rate of 0.1 Hz with 2.3–4.6 nL per pulse. For the AAV1-CAG-floxed-SynGFPrev-WPRE (N = 7 mice), 30 nL were injected, whereas for the pAAV8-FLEX-DTR-GFP (N = 12 mice), 150–300 nL were injected either in one or two locations of the spinal cord. Before and after the pressure injection a waiting time of 10 min was kept. Afterwards, the pipette was retracted, eventual bleeding stopped and the skin sutured. Furthermore, as a control for the DTR experiment, sham injections were performed (N = 7 mice). Using the same method described above, the animals were anesthetized and spinal clamped to a stereotaxic frame. After fat and tissue removal to expose the vertebrae, a glass pipette without virus was lowered to the spinal cord at the Gal+ cells location. The pipette was kept inside the spinal cord for a total of 25 min, after which it was carefully removed, and the animal was sutured. Analgesic (burprenorphine 0.1 mg/kg) was always administered post-surgery. After sufficient time for viral expression (2 weeks for the pAAV8-FLEX-DTR-GFP and 3 weeks for the AAV1-CAG-floxed-SynGFPrev-WPRE) and eventual behavioral experiments (DTR experiment), animals were deeply anesthetized and perfused transcardially with saline, followed by a cold 4% paraformaldehyde solution (PFA) in 0.01 mol/L PBS. Spinal cords were removed from the spine and kept for 1 h in 4% PFA before transferring them for another hour into 0.01 M PBS. Subsequently, spinal cords were stored overnight in 30% sucrose in 0.01 M PBS, 0.1% azide to cryo-protect the tissue. Spinal cords were embedded in frozen section medium and frozen for half an hour at −80 °C in 2-methylbutane solution before mounting them in the cryostat. Spinal cord sections which did not undergo subsequent immunohistochemical staining were cut and mounted on a poly-lysine-coated glass slide at 50 μm, sections with post-hoc immunohistochemical staining were cut and mounted at 30 μm.

Pup viral injections

To infect the BSM-MNs with the light activated channel, channelrhodopsin (ChR2), young animals needed to be used as it has been shown that viral tracers are not able to infect the motor end plate beyond a certain age²⁹. Pups (P3-P6) were briefly separated from their litter and mother and placed in a freezer. After all reflexes were gone, pups were placed on ice and a small incision below the penis was made to access the BSM. Pulled capillaries (length 3 1/2 inches [9 cm]; inner diameter 0.53, outer diameter 1.14 mm; tip diameter 15–20 μm; DrummondScientific, Broomall, PA, USA) were used to inject 1 μl of retroAAV-CAG-hChR2-H134R-tdTomato (28017-AAVrg, Addgene) into the BSM at a rate of 18.4 nl/pulse per second.

After injection, the incision was glued and the pups placed onto a heating pad. Once all reflexes were recovered, the pups were put back to their litter and mother. Injected pups were raised until 2–3 months of age before performing the acute optogenetic experiments as described below.

Electromyogram (EMG) electrodes implantation for chronic recordings

For chronic BSM EMG recording in behaving mice, the animals were implanted with Myomatrix electrodes (model LNT-4×8-BVS-14) kindly made available to us from Sam Sober’s Lab⁸⁷. Mice were anesthetized with 3% isoflurane in oxygen and mouth fixed into a stereotaxic frame (Kopf, Tujunga, CA, USA), allowing the rotation of the body in order to implant the arrays in the bulbospongiosus muscle. During surgery, anesthesia was maintained using 1.5% isoflurane. Three incisions were made to implant the array: first in the pelvic area, then above the animal’s right hind leg and finally a midline incision in the scalp. The skin was carefully separated from the muscle using scissors, using the three incisions to insert the scissors, and finally allowing to subcutaneously guide the arrays’s connector from the pelvic area to the skull to be later secured. Once each thread has been routed subcutaneously and positioned near the BSM, the animal is turned into a supine position and each thread is inserted into the muscle using a suture needle. A suture (size 8–0) was tied to a hole in the array and the needle was then passed through the BSM and used to pull the attached array thread into the muscle. The array was then secured within the muscle by suturing it to the BSM, using additional holes on the array. Afterwards, the incision in the pelvic area was closed and the animal was turned to access the skull. The skull was carefully cleaned and disinfected, and the array’s connector was secured to it using dental cement. If necessary, the incision on the skull was sutured and, after guaranteeing that all the array was under the skin, the incision above the hind leg was also sutured. All mice were single housed post-surgery and used for behavior afterwards.

Nissl stain

The slides were washed 2 times in 0.01 M PBS to remove the excess of frozen section medium. After that, the sections were rehydrated for 40 min in PBS 0.1 M, pH 7.2 (PBS 10x) and permeabilized for 10 min with 0.1% Triton-X (T9284-100ML, Sigma-Aldrich) in PBS 10x. The tissue was washed 2 times 5 min with PBS 10x before incubating them for 20 min with a 1:100 Neurotrace staining solution (in PBS 10x; (NeuroTrace™ 500/525 Green Fluorescent Nissl Stain, N21480, ThermoFisher Scientific; or NeuroTrace™ 530/615 Red Fluorescent Nissl Stain, N21482, ThermoFisher Scientific; or NeuroTrace™ 640/660 Deep-Red Fluorescent Nissl Stain, ThermoFisher Scientific). Subsequently, the tissue was washed with 0.1% Triton-X in PBS 10x for 10 min. After washing 2 times 5 min with PBS 10x, the slides were rinsed with distilled water, dried and coverslipped with Mowiol.

Immunohistochemistry

Immunohistochemical labeling was performed using standard procedures. Briefly, spinal cord sections, which were labeled either for Gastrin releasing peptide (Gastrin Releasing Peptide (GRP) (Porcine) - Antibody, Phoenix Pharmaceuticals, H-027-13, Lot# 01742-1), Substance P (Anti-Substance P Receptor Antibody, Sigma-Aldrich, AB15810, Lot# 3022869), Enkephalin (Anti-Enkephalin/ENK antibody, Abcam, ab85798), CCK (Polyclonal Rabbit anti-Human CCK / Cholecystokinin Antibody, LSBio, LS-C190673, aa26‑33), Osteopontin (Mouse Osteopontin/OPN Antibody, R&D Systems, AF808, Lot# BDO0617401), VGLUT1 (VGLUT 1 antibody, Synaptic Systems, 135 304, Q62362) or Galanin (Anti-Galanin Antibody, Milipore, AB2233, Lot# 3096488), were firstly washed 2 times 5 min with PBS 0.01 M to remove the excess OCT. Afterwards, they were washed 2 times 10 min with PBS 10x and preincubated for 1.5 h at room temperature in a blocking solution (PBS 10x, 1% bovine serum albumin, and 0.3% Triton X-100). Afterwards, primary antibodies were diluted in the same blocking solution at a proportion of 1:100. The primary antibody was incubated on the glass slides overnight at room temperature. Incubation with the primary antibody was followed by 5 times 10 min washing with PBS 10x. Subsequently, we proceeded to detect the primary antibody with a secondary antibody coupled to different fluorophores (Alexa Fluor 488, 594 or 647, Abcam/Thermo Fisher Scientific). The secondary antibody was diluted (1:500) in blocking solution and the reaction was allowed to proceed for 2 h in the dark at room temperature. In some cases, a Nissl stain was performed as described above. After the staining procedure, sections were washed 5 times 10 min with PBS 10x, rinsed with distilled water, dried and cover slipped with Mowiol mounting medium. Immunohistochemical staining for the immediate early gene cFos deferred slightly from the above-described procedure. Namely, after washing with PBS 0.01 M and PBS 10x, the sections were incubated in a different blocking solution (PBS 10x, 0.3% Triton X-100, 4% normal donkey serum, 1% bovine serum albumin) for 1 h at room temperature. The primary antibody (rabbit anti-cFos, Synaptic Systems, 226 003, Rb108B5), diluted 1:500 in blocking solution was added for 2 overnights at 4 °C. Afterwards, the sections were washed 3 times 5 min in PBS 10x, 0.3% Triton X-100, and the secondary antibody (Alexa Fluor 488 or 647, ThermoFisher Scientific) was added in blocking solution, diluted 1:500, for 2 h at room temperature. Finally, the sections were washed 3 times 5 min in PBS 10x, 0.3% Triton X-100, rinsed with distilled water, dried and cover slipped with Mowiol mounting medium. As described before, some slides were also counterstained with Nissl.

Electrophysiology

In vivo electrical stimulations and optogenetic stimulations

For acute electrophysiological experiments, mice (N = 8) and rats (N = 4) were anesthetized by injection of an initial dose of 100 mg/kg ketamine and 7.5 mg/kg xylazine. Respiration, blink and pinch reflex were observed throughout the experiment and, if needed, animals were injected with an extra shot (25%) of ketamine/xylazine mixture or a 25% dose of ketamine alone. The animal’s back, scrotum and right leg were shaved and cleaned with ethanol. Electromyogram electrodes were inserted into the BSM or tibialis anterior (TA) muscles and glued using Vetbond. The back skin was cut along the rostral caudal axis, and the spine was fixed into stereotaxic spinal clamps (Kopf). Muscle and conjunctive tissue were removed before performing a laminectomy along the rostral caudal axis. Spinalization was performed in between the thoracic segments 5 and 6. For electrical stimulations, a 1 MΩ tungsten electrode (World Precision Instruments) was lowered at each microstimulation site (ranging from ~550 to ~950 μm of depth in the dorsal ventral axis, in each spinal cord segment), and currents ranging from 50 to 140 μA were injected using a stimulus isolator (Model no. A365RC, World Precision Instruments) while possible movements were documented. Various stimulation protocols were tested (single, 5 Hz, 50 Hz). Eventual activity in the BSM and TA was monitored via EMG recordings using a custom-made amplifier (×1000) and filtered at 2 kHz. Data acquisition and analysis were performed using spike2 software (CED Cambridge). Tungsten electrodes were moved along the rostral caudal axis of the spinal cord and electrolytic lesions were placed at the location where the biggest EMG responses were encountered. After having mapped the BSM responses along the rostral caudal axis, consecutive electrical stimulations were performed at the spinal cord sites where prominent BSM responses were encountered previously. Inter-stimulus intervals of consecutive electrical stimulations ranged from 15 to 5 min.

In a subset of animals undergoing electrical stimulations (N = 15 mice), sexual behavior (see below) was performed prior to the acute recordings. Animals either ejaculated (N = 8 mice) or performed 5 mounts with intromission (N = 7 mice), before they were anesthetized.

During control experiments, the animals were prepared as above but a waiting time of 1 h was kept before the first stimulation to control that the observed effects are not due to a rundown of the preparation.

For optogenetic stimulations of the Gal+ cells (N = 13 of Gal-ChR2 male mice) and BSM-MNs (N = 10 of C57BL6 mice infected with an AAV expressing ChR2 on the BSM) the experimental procedure was similar except the fact that an optrode (diameter: 1 mm) was moved on top of the spinal cord, in the rostral caudal axis, while monitoring EMG recordings and documenting movements. Likewise, an electrolytic lesion was placed at the position where the light pulses led to the most prominent BSM EMG responses. We tested a variety of stimulation protocols (single, 5 Hz, 10 Hz, 20 Hz, 50 Hz, 100 Hz) and laser powers (5–40 mW).

In vivo juxtacellular recordings of photoidentified cells

Single in vivo juxtacellular recordings were performed as described in ref. ³⁰. Briefly, a glass electrode (resistance ranging from 4 to 6 MΩ) made of borosilicate glass tubes (Hilgenberg) was first lowered at the position where optogenetic stimulations led to the highest activity in the BSM. Pipettes were filled with a ringer solution. Extracellular local field potential recordings were captured while shining the light. Subsequently, cells were searched for by applying a negative current pulse and using an audio monitor (Grass Technologies, AM10) while steps were made in 1.5 μm increments with a micromanipulator (Luigs & Neumann SM-5, Germany) and potential increases in resistance were carefully observed. When spiking activity was detected, electrophysiological recordings were performed in line with optogenetic stimulation protocols (single, 5 Hz, 10 Hz, 20 Hz, 50 Hz) and EMG monitoring. Recordings were amplified (Dagan BVC-700A, Dagan, Minneapolis, MN), low-pass filtered at 10 kHz and sampled at 50 kHz by a data-acquisition interface (Power 1401, CED, Cambridge, England) and controlled and analyzed by the spike2 software (CED, Cambridge, England).

In vitro patching of identified BSM-MNs in new-born male Gal-ChR2 animals

New-born male Gal-ChR2 mice aged postnatal day (P2) to P6 were used in accordance with the guidelines of the French Agriculture and Forestry Ministry for handling animals. The protocol was approved by the local ethics committee of the University of Bordeaux (approval number 2016012716035720). To record specifically from identified MNs innervating the BSM, a crystal of cholera toxin β-subunit conjugated to AlexaFluor 594 (Thermo Fisher Scientific, C34777) was inserted into the BSM with an insect pin 20–24 h before slice preparation procedure in cryo-anesthetized mouse pups. Following the labeling process, spinal cord slices were prepared using the following procedure: mice were anesthetized using isoflurane until all reflexes were gone. After decapitation, the spinal cord was dissected out in an ice-cold sucrose-based saline solution containing the following: 2 mM KCl, 0.5 mM CaCl₂, 7 mM MgCl₂, 1.15 mM NaH₂PO₄, 26 mM NaHCO₃, 11 mM glucose and 205 mM sucrose. The saline was bubbled with 95% O₂, 5% CO₂. Transverse slices (350 μm) of the lower lumbar enlargement and first sacral segments were cut with a vibratome and then transferred to a holding chamber. Slices were allowed to recover in oxygenated aCSF (130 mM NaCl, 3 mM KCl, 2.5 mM CaCl₂, 1.3 mM MgSO₄, 0.58 mM NaH₂PO₄, 25 mM NaHCO₃, 10 mM glucose) for at least 1 h at 30 °C. Whole-cell current-clamp recordings from BSM-MNs, identified by their 594 fluorescence, were made under visual control with a Multiclamp 700B amplifier. Recording glass microelectrodes (4–7 MΩ) were filled with the following: 120 mM K-gluconate, 20 mM KCl, 0.1 mM MgCl₂, 1 mM EGTA, 10 mM HEPES, 0.1 mM CaCl₂, 0.1 mM GTP, 0.2 mM cAMP, 0.1 mM leupeptin, 77 mM d-mannitol and 3 mM Na2-ATP, with a pH of 7.3. All the experiments were performed at room temperature (~23 °C). Data acquisition and analysis were performed using the Axograph software. Experiments were discarded if series resistance increased more than 20% during a given recording period. Polysynaptic transmission was decreased using a high cation solution containing 7.5 mM CaCl₂ and 8 mM MgSO₄ (Liao and Walters, 2002). Throughout recording episodes, GABAergic and glycinergic inputs were blocked with gabazine and strychnine (1 μM each), respectively^88,89. A stimulating optrode connected to an optogenetic laser box (Prizmatrix) was placed above the central canal at the dorsal gray commissure and light pulses were applied at different lengths. Excitatory postsynaptic currents (EPSCs) were recorded from MNs held at −60 mV in current clamp mode. The input resistance of MNs (Rin) was determined from the slope of the voltage-current curve within the linear portion of current traces. AHP parameters were measured after single action potential evoked by short depolarizing current steps (7 ms, 0.25 nA) in current clamp conditions in MNs held at −60 mV by injection of bias current.

Sensory stimulation of the penis and local field potential recordings of photo-identified cells

For sensory stimulation of the penis in parallel with BSM activity monitoring by EMG, the surgical procedure was performed as described above (In vivo electrical stimulations and optogenetic stimulations). To stimulate the penis electrically (by wrapping a nerve cuff electrode around the penis) or mechanically (by locally applying an air puff) the penis was gently pulled out. Electrical (6 V) and mechanical (0.5–1 mbar) stimulations of the penis were done at either 200 Hz (100 pulses) or 5 Hz (3 × 5 pulses, 100 ms). Stimulation protocols were first run in an anesthetized male mouse with an intact brain spinal connection before disrupting the latter by performing a spinalization between the thoracic segments 9 and 10. Electrical and/or mechanical stimulation protocols were then repeated in the spinalized preparation (N = 7 mice).

For local field potential (LFP) recordings of optogenetically tagged Gal+ or BSM-MNs, a glass electrode (resistance ranging from 4 to 6 MΩ) made of borosilicate glass tubes (Hilgenberg) was lowered at the position of the Gal+ cells or the BSM-MNs and optogenetic stimulations were performed until encountering the maximal light triggered LFP response. Subsequently, the glass pipette was left in the place at which the highest light triggered LFP response of BSM-MNs or Gal+ cells was encountered. Finally, the penis and leg were mechanically stimulated with a locally applied airpuff (5 Hz; 5 pulses, 100 ms) while monitoring LFP responses and EMG responses in BSM and TA muscles. Analysis was performed in spike2 and data plotted in Excel and Matlab.

Behavior

Sexual priming prior electrical stimulations

Male mice were single-housed and trained in sexual behavior, with a primed ovariectomized female, until ejaculation was reached in 3 sessions. Afterwards, the animals were divided into two groups: one group was allowed to perform 5 mounts with intromissions (N = 7 mice), and the second group was allowed to perform the whole repertoire of sexual behavior until reaching ejaculation (N = 8 mice). Immediately after the behavior, the animals were anesthetized and used for in vivo electrical stimulations as described above.

cFos experiments

Male Gal-cre x tdTomato mice were single-housed and trained in sexual behavior, with a primed ovariectomized female, until ejaculation was reached in 3 sessions. After one week, the animals underwent a behavioral paradigm and were divided into three different groups: in the first group, animals were allowed to socially interact with a female for 10 min (Aroused, N = 7 mice), without performing mounts or intromissions. Note that the mice were prevented from mounting by putting the hand into the experimental arena thereby separating the mice. The second group of male mice performed five mounts with intromissions (5 Mounts, N = 5 mice), after which the female was removed and in the third group, the animals were allowed to reach ejaculation (Ejaculation, N = 6 mice). In another group of animals (Cage control) male mice were either alone in their home cages, without sexual encounters or female odors present (N = 3 mice), or had the same sexual training but were then alone in the clean behaving box for 10 min (N = 3 mice), in order to assess the baseline neuronal activity in the spinal cord (no differences in the cFos count was observed in these two control conditions). In the last group, animals (N = 5 mice) previously sexually trained, were allowed to interact with another male mouse (5 weeks old) for 10 min or until 10 attack bouts happened. After the behavior session, the experimental mice were placed back in their home cages for 90 min to allow for sufficient cFos expression. Finally, the animals were deeply anesthetized and perfused transcardially with saline, followed by a cold 4% PFA in 0.01 mol/L PBS. Spinal cords were removed from the spine and kept for 1 h in 4% PFA before transferring them for another hour into 0.01 M PBS. Subsequently, spinal cords were stored overnight in 30% sucrose in 0.01 M PBS, 0.1% azide to cryo-protect the tissue. Spinal cords were embedded in frozen section medium for half an hour at −80 °C in 2-methylbutane solution before mounting them in the cryostat. The spinal cords were sectioned at 30μm for post-hoc immunohistochemical staining as described above.

DTR experiments

Male Gal-cre x tdTomato animals (2–3 months old) were single-housed and trained for sexual behavior until ejaculation was reached once. After that, the animals were spinally injected with an AA8V-FLEX-DTR-GFP (N = 12 mice) at the location of the SEG cells (allowing for the specific expression of the Diphtheria Toxin Receptor) or had a sham surgery (N = 7 mice), as described above, and allowed to recover for 2 weeks. Afterwards, the animals underwent another round of sexual behavior training to confirm that they were still reliably ejaculating. After this, the animals received an intraperitoneal injection of 50 ng/g (0.1 mL/10 g) of Diphtheria Toxin (DT, Sigma, D0564-1MG). One week after, the animals were transferred to an experimental box, where they had a 10 min habituation, after which a primed ovariectomized female was introduced. The animals were allowed to perform the full repertoire of sexual behavior and after ejaculation was reached, or 90 min after female introduction passed when sexual behaviors were observed, the animals were returned to their home cage. If the animals did not initiate behavior, the experiment was stopped after 30 min and the sexual experimental paradigm was repeated for two more weeks. In the final session, or after ejaculation, the animals were returned to their home cage and a 90 min interval was kept before perfusion, for cFos expression to occur. Finally, the animals were deeply anesthetized, transcardially perfused and the spinal cords collected for histological processing as described above (cFos experiments).

Erection assessment in DTR experiment

To assess the effect of the Gal+ cells ablation on the erection of DTR animals, a set up with a glass at the bottom was used to observe the pelvic area movements on DTR and sham animals. The set up consists of a mirror at a 45 degrees angle below the experimental box and a glass floor, allowing for the recording of the side view and bottom view of the sexual encounter with the same camera. Male mice underwent the DTR experiments behavioral protocol explained above and had an additional experimental week where they were transferred for the glass set up and the pelvic movements were recorded (Supplementary videos 3 and 4) while the animal performed sexual behavior until ejaculation was reached. Afterwards, the animals were returned to their home cage and a 90 min interval was kept before perfusion, for cFos expression to occur. Finally, the animals were deeply anesthetized, transcardially perfused and the spinal cords collected for histological processing as described above.

In vivo BSM EMG recordings

Male mice were single-housed and trained in sexual behavior, with a primed ovariectomized female, until ejaculation was reached in 3 sessions. Afterwards, the animals were implanted with EMG probes in the BSM as described above. Following a recovery period of 2 weeks post-surgery in their home cages, the animals were put on an experimental box, where they had a 10 min habituation, after which a primed ovariectomized female was introduced. The animals were allowed to perform the full repertoire of sexual behavior until ejaculation was reached, and they were transferred back to their home cage. If the animals did not initiate behavior, the experiment was stopped after 30 min and the sexual experimental paradigm was repeated for at least two more weeks. At the end of the experiment, the animals were sacrificed and the probes recovered to reuse in future experiments. BSM recordings were conducted using an RHD 16-Channel bipolar-input recording headstage (Intan Technologies), connected to an Acquisition Board (Open Ephys), using a sampling rate of 30 KHz. The board and cameras were controlled using a bonsai script (Bonsai Visual Reactive Programming) and the data was band-passed between 850–7,000 Hz and further analyzed using Matlab.

Ovariectomy and hormonal priming

All female mice (N = 20 mice) used as sexual stimuli were ovariectomized. Briefly, female mice were anesthetized using 3% isoflurane in oxygen and placed into a mouthpiece allowing for continuous isoflurane anesthesia. After all reflexes were gone, an incision was made at the center of the lower back. The skin was separated from the muscle towards both sides of the back. After that, on one side, leveled with the hindlimb, the ovary fat pad was located, and a small incision was made in the muscle above this area. The ovary was gently pulled and the connection between the ovary and the uterus was cut using a cauterizer. The same procedure was repeated for the second ovary. Finally, the incision in the back was sutured and the animal was allowed to fully recover on a heating pad. After two weeks of recovery, the animals underwent hormonal priming during which they received an estrogen (1 mg/ml, Sigma E815 in sesame oil) injection 2 days prior to the sexual behavior experiment and a progesterone (5 mg/ml, Sigma P0130 in sesame oil) injection 4 h before the experiment was scheduled. Hormonal priming with estrogen and progesterone was conducted before each experiment.

Quantification and statistical analysis

Histological analysis

After immunohistochemical labeling of cFos (as described above), the spinal cord sections were imaged using a Slide Scanner (Zeiss AxioScan.Z1, Zeiss Microscopy). The obtained images were processed using the Zen Software (Zen 2.6, Zeiss Microscopy), and tiff images of each channel (488 for the cFos, cy3 for the Gal-tdT+ cells and cy5 for the Nissl staining) were exported. These tiff files were, together with the corresponding image of the spinal segment from the mouse spinal cord atlas⁹⁰, opened in Photoshop (Adobe Photoshop, Adobe). Once the overlap between the spinal atlas outlines and the immunohistochemical tiff image was adjusted, the cells present in laminae X (around the central canal and the location of Gal+ cells) of the spinal cord were manually counted by placing dots on each cell. This counting procedure was done for every other slide given the thickness of the slices and thus to prevent double counting. The manual counting procedure was done for both channels, cFos in green and tdtomato in red, by painting circles in Photoshop over the manually scored elements. The Matlab function imfindcircles was used to detect and quantify these marks. To identify overlap between the two channels, the Matlab function knnsearch was used. If two centroids were closer than the radius of a soma, they were considered as co-localized elements. Manual counting of cFos and Gal-tdT+ cells was done blindly to prevent bias. Different experimenters were involved in the counting: ARM counted the animals for the cFos experiment, except for 3 cage alone animals that were counted by CL; LF counted the animals for the Diphtheria Toxin experiment. For the cell normalization presented in Fig. 6, we divided the number of cFos-positive and Gal-tdT+ cells by the total amount of Gal-tdT+ cells.

Behavioral analysis

The DTR, cFos and in vivo BSM recordings behavioral experiments were recorded using two-point gray cameras (Teledyne FLIR), at 60 frames per second. The cameras acquired a top and front video of the cage and were controlled using a bonsai script (Bonsai Visual Reactive Programming). Afterwards, the videos were analyzed using Python Video Annotator (developed at the Champalimaud Foundation). Annotations were conducted in a blind fashion. A wide range of behaviors was annotated, namely: sniffing of the anogenital area of the female; mount attempts (when the male was not able to perform intromissions) mounts with probing (shallow pelvic movements when the male is trying to intromit but still has not inserted the penis); mounts with intra-vaginal thrustings; intra-vaginal thurstings (when the male successfully inserted the penis in the female’s vagina); and finally ejaculation (time between the beginning of shivering and the moment the male dismounted the female). The duration, in frames, for each behavior was then aligned to the camera timestamps, acquired during video recording. Analysis of behavioral timestamp data was performed using a python script in Spyder 3.3.6 (Python).

Statistical analysis

Statistical analysis was performed in Matlab (ttest2, anova2) and Python (scipy and statsmodels). All error ranges represent standard error of the mean. For two-sample comparisons of a single variable, Student’s t test was used, unless in cases when the underlying distributions were non-Gaussian (Shapiro-Wilk test, p < 0.05), where a two-tailed Mann-Whitney-U Test was performed. When multiple variables were compared, a Kruskal Wallis test or a Wilcoxon-Signed-Rank test were used since the data did not follow a Gaussian distribution. Probabilities of the null hypothesis p < 0.05 were judged to be statistically significant. Elements of violin plots: center line, median; box limits, upper (75) and lower (25) quartiles; and whiskers, 1.5x interquartile range. Linear regression was performed in Python using seaborn.regplot, the correlation coefficient and p-value were computed with scipy.stats.pearsonr.

The distribution of Gal-tdT+ cFos+ cells along the L2/L3 spinal segments (Fig. 7G) was performed in Julia (with the KernelDensity package). For each animal, a kernel density estimate of the cell counts along the L2/L3 spinal segments was computed. Then we reported the bootstrapped median and interquartile range (IQR) from 1000 kernel density estimates drawn with replacement from all the animals per condition.

To estimate the inter-thrust interval (ITI) versus mount progression (Fig. 7T), the thrust number was divided by the total number of thrusts in the mount, what we call mount progression. Then data from all the mounts and all the sessions from the different groups were pulled together and a non-parametric kernel regression was performed. The shadow region corresponds to the standard deviation.

A power analysis test was done for the cFos experiment due to the lower number of animals used (5–10 per test group). The G Power software was used with the following parameters: test family—F tests; statistical test—ANOVA: Fixed effects, omnibus, one-way; type of power analysis—Post hoc: Compute achieved power—given α, sample size, and effect size; effect size f—1,457217 (calculated using the means of 0,09082 (cage control), 0,1617 (sexually aroused), 0,1523 (male interaction), 0,1800 (5 mounts), 0,3613 (ejaculation), respective group sizes of 8, 10, 5, 5, 6 and a standard deviation of 0,06092); α error probability—0.05; total sample size—34; number of groups—5. This resulted in a power of 99%, with a noncentrality parameter λ of 6,2500000, a critical F of 2,4674936, a numerator df of 4 and a denominator df of 95.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All data reported in the current manuscript can be found in the following Mendeley data repository https://doi.org/10.17632/xyh6wwd7fz.1. Any further request can be addressed to the corresponding authors (C.L. and S.Q.L.). Source data are provided with this paper.

References

Lenschow, C. & Lima, S. Q. In the mood for sex: neural circuits for reproduction. Curr. Opin. Neurobiol. 60, 155–168 (2020).
Article CAS PubMed Google Scholar
Lucio, R. A. et al. The physiology and ecophysiology of ejaculation. Trop. Subtrop. Agroecosystems 15, S113–S127 (2012).
Google Scholar
Turley, K. R. & Rowland, D. L. Evolving ideas about the male refractory period. BJU Int. 112, 442–452 (2013).
Article PubMed Google Scholar
Valente, S., Marques, T. & Lima, S. Q. No evidence for prolactin’s involvement in the post-ejaculatory refractory period. Commun. Biol. 4, 1–10 (2021).
Article Google Scholar
Allard, J., Truitt, W. A., McKenna, K. E. & Coolen, L. M. Spinal cord control of ejaculation. World J. Urol. 23, 119–126 (2005).
Article PubMed Google Scholar
Giuliano, F. & Clement, P. Neuroanatomy and physiology of ejaculation. Annu. Rev. Sex. Res. 16, 190–216 (2005).
Article PubMed Google Scholar
Corona, G., Jannini, E. A., Vignozzi, L., Rastrelli, G. & Maggi, M. The hormonal control of ejaculation. Nat. Rev. Urol. 9, 508–519 (2012).
Article CAS PubMed Google Scholar
Newman, H. F., Reiss, H. & Northup, J. D. Physical basis of emission, ejaculation, and orgasm in the male. Urology XIX, 341–350 (1982).
Article Google Scholar
Sheu, G., Revenig, L. M. & Hsiao, W. In Men’s Sexual Health and Fertility 13–29 (Springer New York, 2014).
Sachs, B. D. Role of striated penile muscles in penile reflexes, copulation, and induction of pregnancy in the rat. J. Reprod. Fertil. 66, 433–443 (1982).
Article CAS PubMed Google Scholar
Tang, Y., Rampin, O., Giuliano, F. & Ugolini, G. Spinal and brain circuits to motoneurons of the bulbospongiosus muscle: retrograde transneuronal tracing with rabies virus. J. Comp. Neurol. 414, 167–192 (1999).
Article CAS PubMed Google Scholar
Peikert, K., Platzek, I., Bessède, T. & May, C. A. The male bulbospongiosus muscle and its relation to the external anal sphincter. J. Urol. 193, 1433–1440 (2015).
Article PubMed Google Scholar
Breedlove, S. M. Hormonal control of the anatomical specificity of motoneuron-to-muscle innervation in rats. Science 227, 1357–1359 (1985).
Marson, L. & McKenna, K. E. The identification of a brainstem site controlling spinal sexual reflexes in male rats. Brain Res. 515, 303–308 (1990).
Article CAS PubMed Google Scholar
Pescatori, E. S. et al. Electrical stimulation of the dorsal nerve of the penis evokes reflex tonic erections of the penile body and reflex ejaculatory responses in the spinal rat. J. Urol. 149, 627–632 (1993).
Article CAS PubMed Google Scholar
Coolen, L. M., Allard, J., Truitt, W. A. & McKenna, K. E. Central regulation of ejaculation. Physiol. Behav. 83, 203–215 (2004).
Article CAS PubMed Google Scholar
Carro-Juárez, M. & Rodríguez-Manzo, G. The spinal pattern generator for ejaculation. Brain Res. Rev. 58, 106–120 (2008).
Article PubMed Google Scholar
Marson, L. & McKenna, K. E. A role for 5-hydroxytryptamine in descending inhibition of spinal sexual reflexes. Exp. Brain Res. 88, 313–320 (1992).
Article CAS PubMed Google Scholar
Marson, L. & McKenna, K. E. CNS cell groups involved in the control of the ischiocavernosus and bulbospongiosus muscles: A transneuronal tracing study using pseudorabies virus. J. Comp. Neurol. 374, 161–179 (1996).
Article CAS PubMed Google Scholar
Normandin, J. J. & Murphy, A. Z. Serotonergic lesions of the periaqueductal gray, a primary source of serotonin to the nucleus paragigantocellularis, facilitate sexual behavior in male rats. Pharm. Biochem. Behav. 98, 369–375 (2011).
Article CAS Google Scholar
Truitt, W. A. & Coolen, L. M. Identification of a potential ejaculation generator in the spinal cord. Science297, 1566–1569 (2002).
Article ADS CAS PubMed Google Scholar
Dewsbury, D. A. Patterns of copulatory behavior in male mammals. Q Rev. Biol. 47, 1–33 (1972).
Article CAS PubMed Google Scholar
Dewsbury, D. A. Diversity and adaptation in rodent copulatory behavior. Science190, 947–954 (1975).
Article ADS CAS PubMed Google Scholar
Köbbert, C. et al. Current concepts in neuroanatomical tracing. Prog. Neurobiol. 62, 327–351 (2000).
Article PubMed Google Scholar
Misawa, H. et al. Osteopontin is an alpha motor neuron marker in the mouse spinal cord. J. Neurosci. Res. 90, 732–742 (2012).
Article ADS CAS PubMed Google Scholar
Friese, A. et al. Gamma and alpha motor neurons distinguished by expression of transcription factor Err3. Proc. Natl Acad. Sci. USA 106, 13588–13593 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Boyden, E. S., Zhang, F., Bamberg, E., Nagel, G. & Deisseroth, K. Millisecond-timescale, genetically targeted optical control of neural activity. Nat. Neurosci. 8, 1263–1268 (2005).
Article CAS PubMed Google Scholar
Mao, T. et al. Long-range neuronal circuits underlying the interaction between sensory and motor cortex. Neuron 72, 111–123 (2011).
Article CAS PubMed PubMed Central Google Scholar
Stepien, A. E., Tripodi, M. & Arber, S. Monosynaptic rabies virus reveals premotor network organization and synaptic specificity of cholinergic partition cells. Neuron 68, 456–472 (2010).
Article CAS PubMed Google Scholar
Lima, S. Q., Hromádka, T., Znamenskiy, P. & Zador, A. M. PINP: a new method of tagging neuronal populations for identification during in vivo electrophysiological recording. PLoS ONE 4, 1–10 (2009).
Article Google Scholar
Strack, A. M. & Loewy, A. D. Pseudorabies virus: a highly specific transneuronal cell body marker in the sympathetic nervous system. J. Neurosci. 10, 2139–2147 (1990).
Article CAS PubMed PubMed Central Google Scholar
Boldogkoi, Z. et al. Genetically timed, activity-sensor and rainbow transsynaptic viral tools. Nat. Methods 6, 127–130 (2009).
Article CAS PubMed Google Scholar
Saleeba, C., Dempsey, B., Le, S., Goodchild, A. & McMullan, S. A student’s guide to neural circuit tracing. Front Neurosci. 13, 1–20 (2019).
Article Google Scholar
Gong, S. et al. Targeting Cre recombinase to specific neuron populations with bacterial artificial chromosome constructs. J. Neurosci. 27, 9817–9823 (2007).
Article CAS PubMed PubMed Central Google Scholar
Gong, S. et al. A gene expression atlas of the central nervous system based on bacterial artificial chromosomes. Nature 425, 917–925 (2003).
Article ADS CAS PubMed Google Scholar
Madisen, L. et al. A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat. Neurosci. 13, 133–140 (2010).
Article CAS PubMed Google Scholar
Nicholas, A. P., Zhang, X. & Hokfelt, T. An immunohistochemical investigation of the opioid cell column in lamina X of the male rat lumbosacral spinal cord. Neurosci. Lett. 270, 9–12 (1999).
Article CAS PubMed Google Scholar
Coolen, L. M., Veening, J. G., Wells, A. B. & Shipley, M. T. Afferent connections of the parvocellular subparafascicular thalamic nucleus in the rat: evidence for functional subdivisions. J. Comp. Neurol. 463, 132–156 (2003).
Article PubMed Google Scholar
Kozyrev, N., Lehman, M. N. & Coolen, L. M. Activation of gastrin-releasing peptide receptors in the lumbosacral spinal cord is required for ejaculation in male rats. J. Sex. Med. 9, 1303–1318 (2012).
Article CAS PubMed Google Scholar
Sürmeli, G. et al. Molecularly defined circuitry reveals input-output segregation in deep layers of the medial entorhinal cortex. Neuron 88, 1040–1053 (2015).
Article PubMed PubMed Central Google Scholar
Hancock, M. B. & Peveto, C. A. Preganglionic neurons in the sacral spinal cord of the rat: an HRP study. Neurosci. Lett. 11, 1–5 (1979).
Article CAS PubMed Google Scholar
Nadelhaft, I. & McKenna, K. E. Sexual dimorphism in sympathetic preganglionic neurons of the rat hypogastric nerve. J. Comp. Neurol. 256, 308–315 (1987).
Article CAS PubMed Google Scholar
Baron, R. & Jänig, W. Afferent and sympathetic neurons projecting into lumbar visceral nerves of the male rat. J. Comp. Neurol. 314, 429–436 (1991).
Article CAS PubMed Google Scholar
Giuliano, F. et al. Erectile response to hypothalamic stimulation in rats: role of peripheral nerves. Am. J. Physiol. Regul. Integr. Comp. Physiol. 273, 1990–1997 (1997).
Article Google Scholar
Orr, R. & Marson, L. Identification of CNS neurons innervating the rat prostate: a transneuronal tracing study using pseudorabies virus. J. Auton. Nerv. Syst. 72, 4–15 (1998).
Article CAS PubMed Google Scholar
Vizzard, M. A., Erickson, V. L., Card, J. P., Roppolo, J. R. & de Groat, W. C. Transneuronal labeling of neurons in the adult rat brainstem and spinal cord after injection of pseudorabies virus into the urethra. J. Comp. Neurol. 355, 629–640 (1995).
Article CAS PubMed Google Scholar
Marson, L. & Carson, C. Central nervous system innervation of the penis, prostate, and perineal muscles: a transneuronal tracing study. Mol. Urol. 3, 43–50 (1999).
CAS PubMed Google Scholar
Schmidt, M. H. & Schmidt, H. S. The ischiocavernosus and bulbospongiosus muscles in mammalian penile rigidity. Sleep 16, 171–183 (1993).
Article CAS PubMed Google Scholar
Xu, C. et al. Galanin and neurokinin-1 receptor immunoreactive spinal neurons controlling the prostate and the bulbospongiosus muscle identified by transsynaptic labeling in the rat. Neuroscience 134, 1325–1341 (2005).
Article CAS PubMed Google Scholar
Madisen, L. et al. A toolbox of Cre-dependent optogenetic transgenic mice for light-induced activation and silencing. Nat. Neurosci. 15, 793–802 (2012).
Article CAS PubMed PubMed Central Google Scholar
Falgairolle, M. & O’Donovan, M. J. Feedback regulation of locomotion by motoneurons in the vertebrate spinal cord. Curr. Opin. Physiol. 8, 50–55 (2019).
Article PubMed PubMed Central Google Scholar
Lenschow, C. et al. Sexually monomorphic maps and dimorphic responses in rat genital cortex. Curr. Biol. 26, 106–113 (2016).
Article CAS PubMed Google Scholar
Larsson, K. & Sodersten, P. Mating in male rats after section of the dorsal penile nerve. Physiol. Behav. 10, 567–571 (1973).
Article CAS PubMed Google Scholar
Carro-Juárez, M. & Rodríguez-Manzo, G. Role of genital sensory information in the control of the functioning of the spinal generator for ejaculation. Int. J. Impot Res. 17, 114–120 (2005).
Article PubMed Google Scholar
Carro-Juárez, M. & Rodríguez-Manzo, G. Sensory and motor aspects of the coital reflex in the spinal male rat. Behav. Brain Res. 108, 97–103 (2000).
Article PubMed Google Scholar
Borgdorff, A. J., Bernabé, J., Denys, P., Alexandre, L. & Giuliano, F. Ejaculation elicited by microstimulation of lumbar spinothalamic neurons. Eur. Urol. 54, 449–456 (2008).
Article PubMed Google Scholar
McGill, T. E. & Coughlin, R. C. Ejaculatory reflex and luteal activity induction in Mus musculus. J. Reprod. Fertil. 21, 215–220 (1970).
Article CAS PubMed Google Scholar
Soukhova-O’Hare, G. K., Schmidt, M. H., Nozdrachev, A. D. & Gozal, D. A novel mouse model for assessment of male sexual function. Physiol. Behav. 91, 535–543 (2007).
Article PubMed Google Scholar
Krukoff, T. L. In Cell Neurobiology Techniques (eds Boulton, A. A., Baker, G. B. & Bateson, A. N.) 213–230 (Humana Press, 1999).
Azim, E., Jiang, J., Alstermark, B. & Jessell, T. M. Skilled reaching relies on a V2a propriospinal internal copy circuit. Nature 508, 357–363 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
McGill, T. E. Sexual behavior in three inbred strains of mice. Behavior 19, 341–350 (1962).
Article Google Scholar
Frigon, A. & Rossignol, S. Experiments and models of sensorimotor interactions during locomotion. Biol. Cyber. 95, 607–627 (2006).
Article MATH Google Scholar
Guertin, P. A. Central pattern generator for locomotion: anatomical, physiological, and pathophysiological considerations. Front Neurol. 3, 1–15 (2013).
Article Google Scholar
Marson, L., Platt, K. B. & McKenna, K. E. Central nervous system innervation of the penis as revealed by the transneuronal transport of pseudorabies virus. Neuroscience 55, 263–280 (1993).
Article CAS PubMed Google Scholar
Gerendai, I., Tóth, I. E., Boldogkoi, Z., Medveczky, I. & Halász, B. Central nervous system structures labelled from the testis using the transsynaptic viral tracing technique. J. Neuroendocrinol. 12, 1087–1095 (2000).
Article CAS PubMed Google Scholar
Gerendai, I. et al. Identification of neurones of the brain and spinal cord involved in the innervation of the ductus deferens using the viral tracing method. Int J. Androl. 26, 91–100 (2003).
Article PubMed Google Scholar
El-Sakka, A. I. & Lue, T. F. Physiology of penile erection. Sci. World J. 4, 128–134 (2004).
Article CAS Google Scholar
Kolbeck, S. C. & Steers, W. D. Neural regulation of the vas deferens in the rat: An electrophysiological analysis. Am. J. Physiol. Regul. Integr. Comp. Physiol. 263, 331–338 (1992).
Article Google Scholar
Truitt, W. A., Shipley, M. T., Veening, J. G. & Coolen, L. M. Activation of a subset of lumbar spinothalamic neurons after copulatory behavior in male but not female rats. J. Neurosci. 23, 325–331 (2003).
Article CAS PubMed PubMed Central Google Scholar
Schellino, R., Boido, M. & Vercelli, A. The dual nature of Onuf’s nucleus: neuroanatomical features and peculiarities, in health and disease. Front Neuroanat. 14, 1–13 (2020).
Article Google Scholar
Nadelhaft, I., deGroat, W. C. & Morgan, C. Location and morphology of parasympathetic preganglionic neurons in the sacral spinal cord of the cat revealed by retrograde axonal transport of horseradish peroxidase. J. Comp. Neurol. 193, 265–281 (1980).
Article CAS PubMed Google Scholar
Holstege, G. & Tan, J. Supraspinal control of motoneurons innervating the striated muscles of the pelvic floor including urethral and anal sphincters in the cat. Brain 110, 1323–1344 (1987).
Article PubMed Google Scholar
Argiolas, A. & Melis, M. R. Central control of penile erection: role of the paraventricular nucleus of the hypothalamus. Prog. Neurobiol. 76, 1–21 (2005).
Article CAS PubMed Google Scholar
De Groat, W. C. Autonomic Nervous System: Central Urogenital Control. Reference Collection in Neuroscience and Biobehavioral Psychology (Elsevier, 2016).
Argiolas, A. & Melis, M. R. The role of oxytocin and the paraventricular nucleus in the sexual behaviour of male mammals. Physiol. Behav. 83, 309–317 (2004).
Article CAS PubMed Google Scholar
Oti, T. et al. Oxytocin influences male sexual activity via non-synaptic axonal release in the spinal cord. Curr. Biol. 31, 103–114 (2021).
Article CAS PubMed Google Scholar
Nuding, S. C. & Nadelhaft, I. Bilateral projections of the pontine micturition center to the sacral parasympathetic nucleus in the rat. Brain Res. 785, 185–194 (1998).
Article CAS PubMed Google Scholar
Holstege, G., Griffiths, D., De Wall, H. & Dalm, E. Anatomical and physiological observations on supraspinal control of bladder and urethral sphincter muscles in the cat. J. Comp. Neurol. 250, 449–461 (1986).
Article CAS PubMed Google Scholar
Huynh, H. K., Willemsen, A. T. M., Lovick, T. A. & Holstege, G. Pontine control of ejaculation and female orgasm. J. Sex. Med. 10, 3038–3048 (2013).
Article PubMed Google Scholar
Shafik, A. & El-Sibai, O. Mechanism of ejection during ejaculation: identification of a urethrocavernosus reflex. Arch. Androl. 44, 77–83 (2000).
Article CAS PubMed Google Scholar
Tanahashi, M., Karicheti, V., Thor, K. B. & Marson, L. Characterization of bulbospongiosus muscle reflexes activated by urethral distension in male rats. Am. J. Physiol. Regul. Integr. Comp. Physiol. 303, 737–747 (2012).
Article Google Scholar
Núnez, R., Gross, G. H. & Sachs, B. D. Origin and central projections of rat dorsal penile nerve: possible direct projection to autonomic and somatic neurons by primary afferents of nonmuscle origin. J. Comp. Neurol. 247, 417–429 (1986).
Article PubMed Google Scholar
Ueyama, T., Arakawa, H. & Mizuno, N. Central distribution of efferent and afferent components of the pudendal nerve in rat. Anat. Embryol.177, 37–49 (1987).
Article CAS Google Scholar
Pascual, J. I., Insausti, R. & Gonzalo, L. M. Pudendal nerve topography in the rat spinal cord projections studied with the axonal tracer wheat germ agglutimnin conjugated-horseradish peroxidase. J. Urol. 147, 718–722 (1992).
Article CAS PubMed Google Scholar
Qi, L. et al. Krause corpuscles are genital vibrotactile sensors for sexual behaviours. Nature 630, 926–934 (2024).
Article ADS CAS PubMed PubMed Central Google Scholar
Harrison, M. et al. Vertebral landmarks for the identification of spinal cord segments in the mouse. Neuroimage 68, 22–29 (2013).
Article PubMed Google Scholar
Chung, B. et al. Myomatrix arrays for high-definition muscle recording. Elife https://doi.org/10.7554/eLife.88551.2 (2023).
Clarac, F., Pearlstein, E., Pflieger, J. F. & Vinay, L. The in vitro neonatal rat spinal cord preparation: a new insight into mammalian locomotor mechanisms. J. Comp. Physiol. 190, 343–357 (2004).
Article CAS Google Scholar
Taccola, G., Marchetti, C. & Nistri, A. Modulation of rhythmic patterns and cumulative depolarization by group I metabotropic glutamate receptors in the neonatal rat spinal cord in vitro. Eur. J. Neurosci. 19, 533–541 (2004).
Article PubMed Google Scholar
Watson, C., Paxinos, G., Kayalioglu, G. & Heise, C. In The Spinal Cord (eds. Watson, C., Paxinos, G. & Kayalioglu, G. B. T.-T. S. C.) 308–379 (Academic Press, 2009).

Download references

Acknowledgements

We thank B. Roska for reagents (PRV) and much valuable advice, and A. Sathyamurthy for technical assistance (immunohistochemistry protocol for cFos in the spinal cord). We thank the Lima lab for their critical input during this project. We also thank the community at the Champalimaud Research, in particular the Carey Lab for reagents and technical assistance, the ABBE Facility, the Scientific Hardware, Histopathology and Rodent Champalimaud Research Platforms for technical assistances, and Zachary Mainen for reviewing the manuscript. We thank Diana Costa for the figure illustrations used throughout our manuscript. This work was developed with the support from the Champalimaud Foundation, the Institut de Neurosciences Cognitives et Intégratives d’Aquitaine (INCIA), a Human Frontier Science Program Postdoctoral Fellowship (LT000353/2018-L4) (C.L.), a H2020 Marie Skłodowska-Curie Actions Individual Fellowship (799973) (C.L.), a Fundação para a Ciência e a Tecnologia (FCT) PhD Fellowship (PD/BD141576/2018) (A.R.M.), the research infrastructure Congento, co-financed by the Lisboa Regional Operational Programme (Lisboa2020), under the PORTUGAL 2020 Partnership Agreement, through the European Regional Development Fund (ERDF), FCT under the project LISBOA-01-0145-FEDER-022170, an InterEmerging Actions 2020 (301137) (S.Q.L.) and an European Research Council (ERC) Consolidator Grant (772827) (S.Q.L.). We thank the National Institutes of Health (NIH) for their support with the Myomatrix EMG electrodes (NIH grants U24NS126936 and R01NS109237).

Author information

These authors contributed equally: Constanze Lenschow, Ana Rita P. Mendes.

Authors and Affiliations

Champalimaud Research, Champalimaud Foundation, Av. Brasília, Doca de Pedrouços, Lisbon, Portugal
Constanze Lenschow, Ana Rita P. Mendes, Liliana Ferreira, Bertrand Lacoste, Hugo Marques, Nicolas Gutierrez-Castellanos & Susana Q. Lima
INCIA, Université de Bordeaux, CNRS UMR5287, Bordeaux, France
Constanze Lenschow, Camille Quilgars & Sandrine S. Bertrand

Authors

Constanze Lenschow
View author publications
Search author on:PubMed Google Scholar
Ana Rita P. Mendes
View author publications
Search author on:PubMed Google Scholar
Liliana Ferreira
View author publications
Search author on:PubMed Google Scholar
Bertrand Lacoste
View author publications
Search author on:PubMed Google Scholar
Hugo Marques
View author publications
Search author on:PubMed Google Scholar
Nicolas Gutierrez-Castellanos
View author publications
Search author on:PubMed Google Scholar
Camille Quilgars
View author publications
Search author on:PubMed Google Scholar
Sandrine S. Bertrand
View author publications
Search author on:PubMed Google Scholar
Susana Q. Lima
View author publications
Search author on:PubMed Google Scholar

Contributions

C.L., A.R.M., L.F., H.M., C.Q., and S.B. performed experiments. C.L., A.R.M., N.G.C., and B.L. performed all data analysis. C.L. and S.Q.L. supervised the study. C.L., A.R.M., and S.Q.L. wrote the manuscript. S.B. reviewed and edited the final version of the manuscript.

Corresponding authors

Correspondence to Constanze Lenschow or Susana Q. Lima.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Rune Berg and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Files

Supplementary Movie 1

Supplementary Movie 2

Supplementary Movie 3

Supplementary Movie 4

Reporting Summary

Transparent Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Lenschow, C., Mendes, A.R.P., Ferreira, L. et al. A galanin-positive population of lumbar spinal cord neurons modulates sexual arousal and copulatory behavior in male mice. Nat Commun 16, 8282 (2025). https://doi.org/10.1038/s41467-025-63877-2

Download citation

Received: 06 November 2024
Accepted: 01 September 2025
Published: 23 September 2025
Version of record: 23 September 2025
DOI: https://doi.org/10.1038/s41467-025-63877-2