The NS-to-WDR conversion hypothesis and its cracks
The classical framework held that innocuous and noxious circuits are segregated in the dorsal horn under inhibitory interneuron control; after nerve injury, reduced inhibition allows low-threshold mechanosensitive (LTM) afferents to activate nociceptive-specific neurons in superficial laminae, turning touch into pain [1]. Supporting this, in vivo recordings in a mouse cuff model showed that nerve injury increased spontaneously active neurons from 59% to 78%, boosted touch responses in those neurons, and abolished high-threshold (nociceptive-specific) neurons, with spinal GABA-A/glycine blockade mimicking the effect in sham animals [7]. Earlier slice work had already shown that conditioning stimulation could convert previously silent lamina I projection neurons into active ones via glial nitric oxide-mediated presynaptic facilitation, hinting that silent cells were a latent substrate [3]. However, selective silencing or ablation of lamina I NS neurons, including spinothalamic cells, failed to relieve hypersensitivity, and CANE-based silencing of NS neurons did not reduce neuropathic responses to any sensory stimulus, leaving the conversion model unable to explain allodynia [1].
Silent neurons, not converted NS cells, carry the signal
The anchor study adapted an ex vivo parasagittal spinal cord–skin preparation that preserves connectivity from hind paw to lumbar dorsal horn, enabling two-photon Ca2+ imaging of VGLUT2+ excitatory neurons across laminae I–V during natural skin stimulation; 6,795 neurons from 46 mice were analyzed [1]. Under physiological conditions, 63% of responsive neurons were NS, 30% WDR, and 7% LTM, closely matching prior in vivo superficial recordings [1]. After either pharmacological disinhibition (bicuculline plus strychnine) or cuff nerve injury, most excitatory neurons became highly polymodal, and a large number of previously silent neurons were unmasked, responding across a wide dynamic range of modalities in all laminae [1]. Computational modeling that searched for the minimal transformations needed to reproduce the experimental data showed that the large majority of NS neurons remained NS after injury in every lamina, whereas silent→WDR and silent→LTM recruitment accounted for the sensitization [1]. This directly challenges the NS-to-WDR conversion account and instead identifies a dormant circuit as the main driver of touch-evoked pain [1].
How this compares with capsaicin sensitization imaging
A foundational two-photon study using a similar ex vivo skin–cord preparation and CICADA cell profiling found that capsaicin-induced central sensitization produced emergent responses to innocuous input and enlarged receptive fields in superficial dorsal horn neurons, and shifted spinal projection neuron tuning toward low-threshold input [2][8]. That work identified specific excitatory populations (Ex1TACR3/1, Ex2TACR1, Ex5TRHR, Ex6CHRM3) contributing to allodynia and showed that WDR projection neurons were the most responsive to CICADA agonists [2][8]. The new study extends this population-level view by imaging across all laminae, including deep dorsal horn, and by using modeling to distinguish transformation of existing NS cells from recruitment of silent cells; it concludes that silent-neuron activation, not NS conversion, is the primary mechanism [1]. The two datasets are broadly compatible in showing that sensitization expands low-threshold responsiveness, but they differ in the proposed cellular substrate, likely reflecting differences in injury model (capsaicin vs cuff), laminae sampled, and analytical approach [1][2].
Modality specificity and competing mechanistic claims
In vivo calcium imaging after inflammatory or chemotherapy injury has shown modality-specific push–pull shifts in spinal temperature representations: formalin/PGE2 increased heat responses and decreased cold responses via crossover inhibition, while oxaliplatin increased cold and suppressed heat responses [5]. This indicates that injury does not simply unmask a uniform polymodal circuit but can reorganize modality channels in opposite directions, a nuance the silent-neuron framework must accommodate [1][5]. Competing evidence also shows that the contribution of dorsal horn CDK5 to chronic pain depends on the nature of the peripheral injury, with distinct effects on WDR neuron responses to innocuous and noxious stimuli [4]. Together these findings argue against a single universal mechanism and suggest that silent-neuron recruitment may be one of several injury-specific routes to hypersensitivity [1][4][5].
Boundaries of the conclusion and what remains open
The anchor study's conclusions rest on an ex vivo mouse preparation that lacks descending noradrenergic and serotonergic pathways, circulating hormones, immune cells, and GABA-B receptor-mediated inhibition; the authors validated the preparation by matching physiological LTM/NS/WDR proportions to in vivo recordings, but these missing influences could alter silent-neuron behavior in intact animals [1]. The computational model was deliberately restricted to spinal disinhibition and did not incorporate other central sensitization mechanisms, so its predictions about minimal transformations are model-dependent [1]. Translational caution is warranted: neuropathic pain after spinal cord injury in humans is primarily spontaneous rather than evoked, and rodent models have relied heavily on evoked withdrawal reflexes, limiting direct inference from mouse allodynia to human conditions [6]. Clinically, loss of rate-dependent depression of the H-reflex has been proposed as a biomarker of spinal disinhibition in painful diabetic neuropathy, offering a potential bridge, but it does not identify which neurons are recruited [11]. Pharmacological strategies that protect spinal inhibition, such as etifoxine in monoarthritis, reduce pain symptoms and microglial activation in rats, consistent with the idea that disinhibition is a tractable target, but whether they act on silent neurons specifically is unknown [9]. Finally, sensitization of spinal motoneurons by bradykinin complicates interpretation of withdrawal-based pain readouts, since reflex output can be altered independently of sensory circuit changes [10].
About These Sources
This research page is built on 11 peer-reviewed studies — published from 2013 to 2026, 4 from 2024 or later, collectively cited 66 times — selected as the most relevant from 12 studies that passed quality screening, drawn from 72 papers retrieved from a database of over 500 million.
Sources used in this answer
Sensory plasticity of dorsal horn silent neurons as a critical mechanism for neuropathic pain
Using two-photon Ca2+ imaging of 6,795 mouse dorsal horn neurons and computational modeling, this study shows that spinal disinhibition or nerve injury unmasks previously silent neurons that become polymodal or WDR, while most nociceptive-specific neurons remain NS, identifying dormant circuit activation rather than NS-to-WDR conversion as the primary driver of mechanical allodynia.
Cell type-specific calcium imaging of central sensitization in mouse dorsal horn
This foundational two-photon imaging study with CICADA cell profiling shows that capsaicin-induced central sensitization produces emergent low-threshold responses and enlarged receptive fields in superficial dorsal horn neurons and shifts spinal projection neuron tuning toward low-threshold input.
Glial nitric oxide-mediated long-term presynaptic facilitation revealed by optical imaging in rat spinal dorsal horn.
This precursor slice study using voltage-sensitive dye imaging shows that conditioning stimulation can convert previously silent lamina I projection neurons to active ones via glial nitric oxide-mediated presynaptic facilitation, indicating that silent neurons are a latent substrate for sensitization.
The contribution of dorsal horn cyclin‐dependent kinase 5 (CDK5) to the initiation and maintenance of chronic pain is defined by the nature of injury
This competing study reports that the contribution of dorsal horn CDK5 to initiation and maintenance of chronic pain depends on the nature of the peripheral injury, with distinct effects on WDR neuron responses to innocuous and noxious stimuli.
Modality-Specific Modulation of Temperature Representations in the Spinal Cord after Injury.
This validation study using in vivo calcium imaging shows modality-specific push–pull shifts in spinal temperature representations after injury, with inflammation increasing heat and decreasing cold responses via crossover inhibition, and oxaliplatin producing the opposite pattern.
Neuropathic Pain After Spinal Cord Injury: Challenges and Research Perspectives.
This limitation review notes that human spinal cord injury neuropathic pain is primarily spontaneous rather than evoked, while rodent models rely on evoked withdrawals, and that pain responses can arise from pathological changes anywhere along the sensory neuraxis.
Loss of inhibitory tone on spinal cord dorsal horn spontaneously and nonspontaneously active neurons in a mouse model of neuropathic pain.
This precursor in vivo recording study in a mouse cuff model shows that nerve injury increases spontaneously active dorsal horn neurons from 59% to 78%, enhances touch responses, abolishes high-threshold neurons, and that GABA-A/glycine blockade mimics these changes in sham animals.
Cell type-specific calcium imaging of central sensitization in mouse dorsal horn.
This validation study provides a population-level view of central sensitization, showing that capsaicin-sensitized dorsal horn neurons develop emergent low-threshold responses and that spinal projection neurons shift tuning toward low-threshold input.
Etifoxine analgesia in experimental monoarthritis: a combined action that protects spinal inhibition and limits central inflammatory processes.
This limitation study in a rat monoarthritis model shows that etifoxine reduces pain symptoms, microglial activation, and proinflammatory mediators while amplifying spinal GABAergic inhibition and preventing glycinergic disinhibition.
Sensitization of neonatal rat lumbar motoneuron by the inflammatory pain mediator bradykinin.
This validation study shows that bradykinin sensitizes spinal motoneurons via B2 receptor, phospholipase C, InsP3, and calmodulin signaling, increasing self-sustained spiking and questioning the use of withdrawal reflexes as a surrogate pain model.
The H-Reflex as a Biomarker for Spinal Disinhibition in Painful Diabetic Neuropathy.
This limitation review summarizes evidence that impaired rate-dependent depression of the spinal H-reflex occurs in animal models of diabetes and in diabetic patients with painful versus painless neuropathy, proposing it as a clinical biomarker of spinal disinhibition.