Single-nuclear RNA sequencing reveals an ATF3-independent sensory-neuron program after surgical incision

Abstract Background Acute postoperative pain is common and often treated with opioids, but which sensory-neuron changes are responsible for the peripheral drive of pain is poorly defined. Skin incision induces activating transcription factor 3 (ATF3), the canonical marker of nerve injury, in dorsal root ganglion (DRG) neurons, and ATF3 marks the neurons that remain hyperexcitable. Whether ATF3 is required for postoperative pain is unknown. Methods Adult mice of both sexes underwent hind-paw plantar incision. Postoperative nociceptive behavior was compared between sensory-neuron–specific ATF3 conditional-knockout mice ( Avil Cre/+ ; Atf3 fl/fl ) and littermate controls using assays that probe distinct primary-afferent modalities (von Frey; Hargreaves; dynamic light touch) together with an operator-independent index of spontaneous injury behavior. Single-nucleus RNA sequencing of wild-type and ATF3-null DRG sensory neurons (naïve and postoperative day 1) characterized the initiating injury program; transcription-factor–activity inference, RNAscope, and phospho–c-Jun immunostaining examined c-Jun; and the oral dual leucine zipper kinase (DLK) inhibitor GNE-3511, given at the time of incision, was tested behaviorally. Results Incision transiently induced ATF3 in a subset of DRG neurons, returning to baseline by day 14, in parallel with the two-week course of nociceptive behavior. Deleting ATF3 in sensory neurons did not change the onset or resolution of nociceptive behavior on any readout, in either sex, across mechanical, thermal, and tactile modalities. Single-nucleus profiling showed a broad surgery-responsive nociceptor program in which ATF3 was the most strongly induced gene, yet only 23% of surgery-responsive neurons expressed it. Without ATF3, the program was remodeled but not abolished, and transcription-factor–activity inference nominated c-Jun as a candidate ATF3-independent factor; incision induced c-Jun in vivo. Systemic inhibition of DLK, which activates c-Jun pathway, lowered c-Jun phosphorylation and reduced evoked nociceptive hypersensitivity. Conclusions Our results indicate that ATF3 marks part of the injured sensory neuron but does not drive acute post-surgical nociceptive behavior, and that it is not required for the postsurgical pain behaviors across afferent modalities carried by molecularly distinct nociceptor classes. The underlying injury program is broader than ATF3 and is nominated to depend on c-Jun. Because DLK inhibition, given at the time of incision, reduced evoked hypersensitivity, we propose the DLK–JNK–c-Jun axis as a candidate perioperative non-opioid target, which will require further genetic validation.