Mouse pain models recruit overlapping brain regions but split into different network patterns
Inflammatory and neuropathic pain can look similar at the bedside: persistent hypersensitivity, altered affect, and the burden of living with ongoing pain. In a mouse study comparing complete Freund’s adjuvant and spared nerve injury, the central nervous system response was neither uniform nor random. Both models activated many of the same broad brain systems, yet the regional organization of those responses diverged in ways that point to distinct circuit architectures for different pain states.
The work set out to answer a straightforward but difficult question in pain neuroscience: do different peripheral injuries converge on a shared brain-wide activation map, or does each model leave a different neuronal signature across the brain? Using c-Fos immunohistochemistry as a marker of recent neuronal activation, investigators surveyed 50 subregions across seven major brain divisions, then aligned the images to the Allen Mouse Brain Atlas for region-by-region comparison. The design allowed a broad look at how nociceptive input, emotional processing, and motivational circuitry are recruited when pain becomes persistent.
Both mouse models developed persistent nociceptive hypersensitivity. Both also showed anxiety-like behavior. No detectable depression-like behavior emerged in either group. Those behavioral readouts matter because chronic pain is rarely confined to sensory discomfort alone; it often reaches into threat appraisal, arousal, and avoidance behavior, which are functions distributed across thalamic, hypothalamic, cortical, and limbic nodes.
Shared activation was extensive, but the counts were not identical
Regional c-Fos increases appeared across a large portion of the brain in both models. CFA increased c-Fos expression in 28 regions, while SNI increased expression in 34 regions. Twenty regions were commonly activated across the two conditions. These shared areas extended across sensory, affective, and motivational systems, suggesting that different pain etiologies still converge on a core pain-processing backbone.
The overlap is biologically plausible. Persistent nociceptive drive does not stop at spinal relay stations; it engages ascending sensory pathways, evaluative circuits, and state-control networks that shape vigilance, aversion, and coping. In practical terms, this means an inflammatory insult and a nerve injury can end up speaking to the brain in similar language, even if the upstream triggers are different.
The common regions also argue against a simplistic view that one pain type lives only in a sensory pathway and another only in affective circuitry. Instead, the data support a more integrated picture in which sensory-discriminative and emotional-motivational domains are recruited together. Pain perception, after all, is built from both the intensity of the input and the brain’s interpretation of what that input means.
Thalamic and hypothalamic hubs dominated inflammatory pain
The CFA network showed relative prominence of thalamic and hypothalamic hubs. Two regions stood out in particular: the paraventricular thalamus and the dorsomedial hypothalamus. Their appearance as central nodes fits with the role of these structures in arousal, stress responsivity, autonomic regulation, and the integration of nociceptive salience.
This pattern suggests that inflammatory pain may lean more heavily on diencephalic control centers that coordinate broad state changes. The brain does not experience sustained inflammation as a purely local sensory event. It registers a threat that can alter vigilance, autonomic tone, and behavioral prioritization, and thalamic-hypothalamic coordination is well suited to that task.
One sentence captures the contrast: inflammatory pain in this dataset concentrated around hub-like diencephalic control points.
Neuropathic pain spread across sensory and corticolimbic circuitry
SNI, by contrast, produced a more distributed profile with sensory and corticolimbic hub prominence. The primary somatosensory cortex, secondary motor cortex, central amygdala, and hippocampal CA3 emerged as important nodes. Rather than clustering around a small number of dominant diencephalic hubs, the network appeared more dispersed across regions tied to sensory mapping, motor output, emotional learning, and memory-related processing.
That architecture is consistent with neuropathic pain’s tendency to reorganize perception and behavior over time. Injury to a nerve can alter incoming sensory signals and prompt broader plastic changes in how those signals are encoded, anticipated, and avoided. The amygdala and hippocampus are especially relevant here because chronic pain frequently interacts with fear learning, contextual memory, and anticipatory stress. A distributed corticolimbic pattern may therefore reflect a brain trying to continuously interpret an aberrant and persistent signal.
The motor cortex signal is also notable. Pain does not merely produce sensation; it changes movement, guarding, and action selection. Recruitment of motor-related cortex may reflect the brain’s effort to adapt behavior around a persistent threat.
Understudied nuclei appeared in both models
Several comparatively understudied regions showed model-associated activation, including the dorsal endopiriform, lateral septal, and basomedial amygdalar nuclei. These findings matter because chronic pain research still gravitates toward a familiar set of canonical nodes, while less examined areas may participate in coordination, modulation, or context-specific valuation of painful experience.
A single additional sentence deserves emphasis: pain circuitry extends beyond the usual suspects.
These lesser-studied nuclei could be valuable entry points for later mechanistic work. Their appearance in both inflammatory and neuropathic pain models raises the possibility that they contribute to shared motivational or affective processing, even if their exact roles remain undefined in this study. Because the analysis was based on c-Fos activation, the data identify candidates rather than prove functional necessity, but candidate discovery is precisely where circuit mapping begins.
What c-Fos covariance can and cannot tell us
The investigators also used interregional c-Fos covariance analysis to assess co-activation patterns and infer candidate network hubs. This approach adds a systems-level layer to the single-region counts by asking which areas tend to vary together across animals. The resulting patterns showed that shared activation does not mean shared organization. CFA and SNI both lit up overlapping territories, but their co-activation architecture differed, implying that the brain may assemble pain-related activity into distinct network configurations depending on the initiating insult.
That distinction has direct experimental value. Descriptive activation maps can guide future work toward circuit dissection, optogenetic manipulation, or cell-type-specific recording, especially when a region appears repeatedly in one model but not the other. The study does not establish directed connectivity, causality, or a therapeutic target on its own. It does, however, narrow the field to specific hubs and assemblies worth testing in mechanistic experiments.
For translational pain research, the message is clear enough without overstatement: inflammatory and neuropathic pain share a substantial brain-wide footprint, yet they do not organize that footprint in the same way. If future studies confirm that these hub patterns map onto distinct behavioral or autonomic phenotypes, treatment strategies may eventually need to account for circuit-level differences rather than treating chronic pain as a single central process.