Deconstruction of a spino-brain–spinal cord circuit that drives chronic pain
Article Meta data
Article Date: 1 April 2026
Article URL: https://www.nature.com/articles/s41586-026-10296-y
Article Title: Deconstruction of a spino-brain–spinal cord circuit that drives chronic pain
Article Image: (not provided)
Summary
This Nature paper maps a long-range multisynaptic loop that is necessary and sufficient for chronic mechanical pain after nerve injury or inflammation. Starting from genetically and virally targeted OPRM1+ RVM→spinal cord (RVM SC) neurons, the authors trace an ascending spinothalamic route to the somatosensory cortex (SS) and a descending return via the lateral superior colliculus (lSuC) and RVM back to the spinal cord. They show OPRM1+ RVM SC neurons are activated by noxious input, are sensitised after injury, and are specifically required to initiate and maintain mechanical hypersensitisation and cold allodynia — but have minimal role in acute protective nociception. Silencing or ablating nodes in this loop (SS→lSuC→RVM→spinal cord or the STT→thalamus→SS pathways) eliminates chronic mechanical pain, while repeated activation at any node induces persistent hypersensitivity in healthy mice. By contrast, PAG inputs are not the main driver of this sensitisation, and tactile discrimination is routed separately via dorsal column nucleus (DCN)→VPL.
Key Points
- OPRM1+ RVM SC neurons (mostly GABAergic) are nociception-responsive and sensitised after peripheral injury; DAMGO inhibits them, confirming functional μ-opioid receptors.
- A long-range loop (spinal cord → VPL/Po thalamus → SS (layer 5) → lateral superior colliculus (lSuC) → RVM → spinal cord) is necessary and sufficient for chronic mechanical hypersensitisation.
- Silencing or ablating OPRM1+ RVM SC neurons prevents initiation and reverses maintenance of neuropathic and inflammatory mechanical hypersensitivity and associated affective pain behaviours.
- Excitatory lSuC→RVM input (predominantly VGLUT2+) is a key driver of mechanical hypersensitisation; PAG→RVM excitatory input is not required for this effect.
- Two parallel spinothalamic pathways (STT→VPL→SS and STT→Po→SS) can each drive chronic pain; silencing both abolishes SNI-induced mechanical pain, while neither alone is essential.
- DCN→VPL is required for tactile discrimination but not for injury-induced pain sensitisation — ascending pathways are functionally segregated.
- Repeated activation of nodes in this circuit produces long-lasting hypersensitivity and anxiety-like states; circuit nodes are potential selective targets for chronic pain therapy that spare acute protective nociception.
Content summary
The authors created an Oprm1-Cre knock-in mouse and improved retrograde AAV tools (AAV8-retro) to target spinally projecting RVM neurons that express OPRM1. They show most OPRM1+ RVM SC neurons are GABAergic, project via the lateral funiculus to dorsal horn superficial laminae, and are activated by noxious mechanical and cold stimuli. After spared nerve injury (SNI) or CFA inflammation these neurons become hyper-responsive to mechanical and cold stimuli (but not heat).
Functional experiments used chemogenetic activation, silencing and caspase-mediated ablation: acute manipulation had little effect on baseline nociception, but ablation before injury prevented development of mechanical hypersensitivity, and silencing after injury reversed established neuropathic pain, reduced spontaneous pain behaviours and normalised brain-wide FOS responses. Repetitive activation induced persistent mechanical hypersensitivity in otherwise healthy mice.
Monosynaptic tracing (cTRIO) identified strong excitatory input from the lateral superior colliculus (lSuC) to OPRM1+ RVM SC neurons; silencing the lSuC→RVM pathway blocked mechanical and inflammatory mechanical sensitisation and pain-induced aversion, while repetitive activation induced chronic pain. Upstream, layer 5 somatosensory cortex (SS) projections to lSuC (but not anterior motor cortex projections) are necessary and sufficient for driving the lSuC→RVM pathway and resulting chronic mechanical pain. Ascending spinothalamic inputs via VPL and Po relay to SS; either STT→VPL→SS or STT→Po→SS is sufficient (redundant) to trigger the downstream loop. DCN→VPL inputs selectively support tactile discrimination rather than injury sensitisation.
The authors propose a model where injury-evoked STT activity engages thalamocortical circuits that drive SS→lSuC→RVM recruitment, and OPRM1+ inhibitory descending RVM SC neurons may disinhibit a spinal ‘gate’ interneuron population, permitting non-painful input to access STT output and produce pathological pain. Identifying those local spinal interneurons is the next step.
Context and relevance
This work reframes chronic mechanical pain as a self-sustaining spino–brain–spinal loop that can be switched on or off by defined long-range nodes. It separates pathways for tactile discrimination (DCN→VPL) from those that produce injury-induced sensitisation (STT→thalamus→SS→lSuC→RVM→spinal cord), clarifying why some interventions (for example, PAG stimulation) produce analgesia but do not prevent sensitisation. For researchers and clinicians, the paper highlights several intervention points (SS, lSuC, OPRM1+ RVM SC neurons, STT→thalamus) where targeted modulation could relieve chronic pain while preserving acute protective nociception.
Author style
Punchy: This is a clear, high-impact mechanistic dissection. The team uses modern genetic, tracing and chemogenetic tools to move from cell type to circuit to behaviour — and the circuit they identify is both necessary and sufficient for chronic mechanical pain. If you work on pain mechanisms or translational neurotherapeutics, the details matter — read the full Methods and tracing data.
Why should I read this?
Quick take: want to know how acute injury becomes persistent pain? This paper maps a specific loop from spinal cord up to somatosensory cortex and back that actually drives chronic mechanical sensitisation. It points to targets that might treat chronic pain without killing protective pain — so yes, it’s worth a skim or a deep dive depending on whether you’re a clinician, translational scientist or basic neurobiologist. We saved you the slog: the main circuit and the nodes that work are spelled out.