Hypothesis / aims of study
Lower urinary tract function (LUTF) relies on the seamless coordination of autonomic and somatic pathways to control urine storage and voiding. The spinal cord is the critical hub for integrating sensory inputs and generating motor outputs essential to bladder control. Yet, the specific spinal neuronal populations that mediate these functions remain largely undefined. Previous efforts have outlined the general locations of neurons regulating LUTF in the spinal cord, but their cellular identities and molecular characteristics have remained unclear. In this original work, we aimed to generate a comprehensive molecular and spatial atlas of the mouse lumbosacral spinal cord to 1) define transcriptionally distinct neuronal populations, 2) determine their anatomical organization, and 3) identify candidate cell types involved in lower urinary tract control.
Study design, materials and methods
Adult Fos-TRAP2::H2B-Trap transgenic mice of both sexes were implanted with a catheter in the bladder dome. The catheter was tunneled to the back of the neck and connected to an infusion harness. After a one-week recovery period, saline was infused into the bladder of awake mice at 20 μL/min for 2 hours. 4-Hydroxytamoxifen (4-OHT; 50 mg/kg, i.p.) was administered, and infusion was continued for an additional 2 hours to induce recombination and permanently label activated neurons. Spinal cords were collected one week later, and lumbosacral segments (L6–S2) from TRAP and wild-type (WT) mice were processed for single-nucleus RNA sequencing (snRNA-seq). Single-nuclei suspensions were generated by “Dounce” homogenization followed by removal of myelin and debris. Fluorescence-activated nuclei sorting was used to isolate labeled (TRAP) and total (WT) nuclei. Libraries were prepared using the 10x Genomics Chromium 3′ GEM-X platform. Furthermore, high-resolution spatial transcriptomics was performed using the 10x Genomics Visium HD platform on lumbosacral spinal cord cross-sections (n=18; L6–S1). The spatial and snRNA-seq datasets were integrated to assign transcriptionally defined populations to anatomical regions. Promising candidate marker genes were validated in human sacral spinal cord tissue (NIH NeuroBioBank) using RNAscope.
Results
The combined analysis of TRAP-enriched and wild-type samples yielded a dataset of 24,808 nuclei, identifying 34 transcriptionally distinct neuronal populations in the lumbosacral spinal cord. Based on canonical marker gene expression, these populations were classified into excitatory (Slc17a6⁺), inhibitory (Gad2⁺), and cholinergic motor neurons (Chat⁺). Within these groups, further molecular characterization revealed heterogeneity, consistent with the presence of locally projecting interneurons (Nfib⁺) and long-range projection neurons (Zfhx3⁺) (3). No clear sex-specific clustering was observed at the current resolution.
Analysis of TRAP versus wild-type contributions across clusters revealed differential enrichment of specific populations following bladder activation, thereby identifying neuronal populations potentially involved in LUTF.
Spatial transcriptomics mapped transcriptionally defined populations to distinct anatomical regions, including the superficial dorsal horn, intermediolateral cell column, and dorsal gray commissure, and enabled characterization of non-neuronal cell types within the local cellular environment (Figure 1).
Interpretation of results
Integration of activity-dependent labeling with snRNA-seq and spatial transcriptomics facilitated the detailed characterization of neuronal populations in the lumbosacral spinal cord and their involvement in LUTF. Enrichment of specific clusters in the TRAP samples supports their recruitment during bladder activity. Localizing these populations to regions implicated in LUTF, including the dorsal gray commissure, intermediolateral cell column, dorsal horn, and Onuf’s nucleus, further strengthens their functional relevance. Additionally, the combined molecular and spatial resolution allows inference of interactions between neuronal populations and highlights candidate molecular targets for therapeutic intervention in lower urinary tract dysfunction.