CtBP1–LSD1 complex drives ErbB2 activation via H3K9me2 demethylation in DRGs during paclitaxel-induced neuropathic pain Authors: Hsien-Yu Peng, Cheng-Yuan Lai, Gin-Den Chen, Yun-Chih Peng, Chou-Ming Yeh, Tzer-Bin Lin, Hsueh-Hsiao Wang, Jen-Kun Cheng, Po-Sheng Yang, Chia-Man Chou, Dylan Chou, Kuan-Hung Lin, Siao-Tong Nie, Ming-Chun Hsieh Published: 2025 ABSTRACT AbstractPaclitaxel (PTX), a commonly utilized chemotherapy drug, is linked to peripheral neuropathy, which limits dosing and significantly affects patients' quality of life. C-terminal binding protein 1 (CtBP1) is a transcriptional coregulator that participates in epigenetic gene regulation, but its role in PTX-induced neuropathic pain remains unclear. In this study, the role of CtBP1 in PTX-induced neuropathic pain is examined, with a focus on its epigenetic regulation in the dorsal root ganglia (DRGs). PTX administration markedly increased CtBP1 protein levels in DRG neurons, which coincided with the development and continuation of mechanical allodynia and thermal hyperalgesia in rat models. Our findings also revealed that CtBP1 interacts with the histone demethylase LSD1—a regulator of H3K9me2—at ErbB2 promoter sites in DRG neurons. PTX treatment increased CtBP1 protein levels, which subsequently induced LSD1 expression and decreased H3K9me2 protein levels at the ErbB2 promoter, indicating epigenetic activation of ErbB2 signaling in DRG neurons implicated in neuropathic pain. Reducing either CtBP1 or LSD1 expression reversed ErbB2 upregulation and attenuated PTX-induced pain sensitivity. These results suggest that the CtBP1-LSD1 complex epigenetically increases ErbB2 expression in DRG neurons, contributing to PTX-induced neuropathy. Targeting the CtBP1–LSD1 pathway could represent a promising therapeutic strategy for the treatment of chemotherapy-induced neuropathic pain.Graphical AbstractGraphical Headlights• CtBP1–LSD1–ErbB2 signaling mediates paclitaxel-induced neuropathic pain in DRG neurons.• Paclitaxel upregulates CtBP1, which interacts with LSD1 to demethylate histone H3K9 at the ErbB2 promoter.• This epigenetic modification activates ErbB2 transcription, contributing to neuropathic pain development. Supplementary InformationThe online version contains supplementary material available at 10.1007/s10565-025-10122-7.Keywords: Paclitaxel, CtBP1, LSD1, ErbB2, DRG, Neuropathic pain Introduction Neuropathic pain is a chronic condition arising from lesions or diseases affecting the somatosensory system, often refractory to conventional analgesics and severely impairing patients’ quality of life (Finnerup et al. 2021). Among its various forms, chemotherapy-induced neuropathic pain (CINP) is a major dose-limiting adverse effect of anticancer agents, including platinum compounds, vincristine, and taxanes. Paclitaxel (PTX), a widely used taxane-based chemotherapeutic agent for breast, ovarian, and lung cancers, frequently induces peripheral neuropathy characterized by mechanical allodynia, hyperalgesia, and sensory loss, which can persist long after treatment cessation (Dong Yang et al. 2024). Despite extensive study, the molecular mechanisms underlying PTX-induced neuropathic pain remain unclear. Understanding the molecular alterations within the dorsal root ganglion (DRG)—a key site of sensory processing—is essential for developing effective therapies. Human epidermal growth factor receptor 2 (ErbB2/HER2) is a receptor tyrosine kinase that regulates cell growth and survival through heterodimerization with other ErbB family members. Overexpression of ErbB2 drives tumor progression and invasion and is linked to poor prognosis (Di Lisa et al. 2024). In the nervous system, ErbB2 also modulates neuronal activity and pain signaling (Schlam and Swain 2021). Following peripheral nerve injury, ErbB2 expression increases in ipsilateral DRGs, and its inhibition alleviates neuropathic pain (Mizobuchi et al. 2013). Moreover, PTX alters epigenetic marks at the ErbB2 promoter in cancer cells (Mo et al. 2017), suggesting that aberrant ErbB2 activation in DRGs may contribute to PTX-induced neuropathy. C-terminal binding proteins (CtBPs) are transcriptional coregulators that control gene expression via chromatin remodeling. The two isoforms, CtBP1 and CtBP2, are implicated in tumor progression and invasion, as well as with unfavorable clinical outcomes (Raicu et al. 2021). Recent studies have highlighted the role of CtBPs in chemotherapy sensitivity (Hao et al. 2019; He et al. 2021). CtBP1 deficiency sensitizes tumor cells to PTX and reduces metastasis (Akhtar et al. 2023). Notably, CtBP1-containing complexes can increase NF-κB-dependent transcriptional activity (Lou et al. 2023) and modulate oxidative stress responses (Wang et al. 2020). PTX has been found to induce inflammatory and oxidative responses in the peripheral nervous system, such as through NF-κB pathways (Naziroglu and Braidy 2017). Furthermore, our recent findings further demonstrate that CtBP1-dependent epigenetic modulation of DRG neurons contributes to neuropathic pain following nerve injury (Lai et al. 2024). These insights raise the intriguing possibility that CtBP1-related pathways in the DRG may also participate in PTX-induced neuropathic pain. In addition, although CtBP proteins are traditionally viewed as transcriptional corepressors, emerging evidence indicates that they can also act as transactivators for specific genes (Ray et al. 2014). Notably, the expression of the long noncoding RNA CtBP1-AS is positively correlated with tumor size, histological grade, and ErbB2 levels in human cancers (Cui and Geng 2019), suggesting a potential regulatory connection between CtBP1 and ErbB2. Therefore, exploring whether CtBP1 influences the epigenetic state of the ErbB2 promoter, potentially leading to upregulation in DRG neurons during PTX-induced neuropathic pain, is critical. While CtBP proteins are classically recognized as transcriptional corepressors, emerging evidence suggests that they can also function as coactivators, depending on the chromatin context and interacting partners (Ray et al. 2014). Because CtBPs lack direct DNA-binding ability, they rely on transcription factors to recruit chromatin modifiers such as HDACs, G9a, and lysine-specific demethylase 1 (LSD1) to gene promoters (Dcona et al. 2017; Lim et al. 2024). LSD1 (also known as KDM1A) is a flavin-dependent histone demethylase that modulates gene expression by removing methyl groups from histone H3 on lysine 4 (H3K4me2) or lysine 9 (H3K9me2), depending on its associated complex (Garcia-Bassets et al. 2007). Importantly, CtBP1 and LSD1 have been identified as coactivators that jointly reduce H3K9me2 repressive marks to promote transcriptional activation (Ferrarese et al. 2023) (Ray et al. 2014). Moreover, LSD1 directly regulates ErbB2 expression through H3K9me2 demethylation (Lin et al. 2021). Knockdown of LSD1 decreases its occupancy at the ErbB2 promoter, leading to increased H3K9me2 expression and reduced ErbB2 transcription (Lim et al. 2010). In the nervous system, the expression of LSD1 is upregulated in injured DRGs, contributing to heightened pain hypersensitivity associated with neuropathic pain (Wei et al. 2022). Similarly, paclitaxel treatment increases LSD1 expression in cancer cells, further implicating LSD1 in PTX-induced molecular alterations (Chen et al. 2024c).. Taken together, these findings suggest a potential CtBP1–LSD1–ErbB2 regulatory axis, wherein CtBP1-LSD1 complexes may act epigenetically to increase ErbB2 transcription in DRGs under PTX exposure. Materials and methods Experimental animals Adult male Sprague–Dawley rats (200–250 g; mean age, 7 weeks) were primarily used in this study. For gender-difference experiments, adult female Sprague–Dawley rats (200–250 g; mean age, 7 weeks) were additionally included. All rats were obtained from BioLASCO Taiwan Co., Ltd (Taipei, Taiwan). The animals were housed in groups of three to four per cage under standard laboratory conditions (temperature, 22 ± 2 °C; humidity, 55 ± 10%; 12-h light/dark cycle), with free access to food and water. To minimize selection bias, the Research Randomizer tool (https://www.randomizer.org/) was used to assign animals randomly to experimental groups. Experimenters were blinded to the treatment assignments throughout all procedures. For behavioral experiments, five or six rats per group were used, while five rats per group were used for biochemical and histological analyses. In experiments involving pain models (paclitaxel-induced neuropathic pain), animals were housed individually or in pairs after surgery to prevent aggressive behavior and minimize potential interference with pain assessment. All animal care and experimental procedures were approved by the Institutional Animal Care and Use Committee of Mackay Medical College, New Taipei, Taiwan (Approval No. A111018, A11320028), and were conducted in accordance with the Guide for the Care and Use of Laboratory Animals (National Research Council, USA). Animal model of paclitaxel-induced neuropathic pain Neuropathic pain was induced using a well-established paclitaxel (PTX)-induced model, as described in previous studies (Abd-Elrazek et al. 2020; Duggett et al. 2017; Yeh et al. 2024). PTX (Taxol; Tocris Bioscience, Bristol, UK) was provided at a concentration of 6 mg/mL in a 1:1 mixture of Cremophor EL and ethanol, then diluted with sterile saline to 1 mg/mL for administration. The rats received intraperitoneal injections of 2 mg/kg on days 0, 2, 4, and 6, totaling a dose of 8 mg/kg. Control rats were given a vehicle solution made of diluted Cremophor EL/ethanol in saline, mirroring the PTX formulation. Behavioral assessments All behavioral assessments were conducted by investigators blinded to the treatment conditions. Mechanical allodynia: Mechanical hypersensitivity was evaluated using calibrated von Frey filaments (Stoelting Co., Wood Dale, IL, USA). Before testing, rats were placed individually in transparent acrylic enclosures positioned on an elevated wire mesh platform and allowed to acclimate for 30 min. The paw-withdrawal threshold in response to punctate mechanical stimulation was measured using filaments with logarithmically increasing stiffness ranging from 0.08 to 26 g. Testing was performed on both hind paws to evaluate mechanical sensitivity. The assessment began with a 4 g filament, which was applied perpendicularly to the plantar surface of the paw for approximately 6 s, ensuring the filament bent slightly upon contact. A brisk withdrawal or paw flinching was recorded as a positive response. If a positive response occurred, the next lower-force filament was applied; otherwise, the next stiffer filament was used. The 50% paw-withdrawal threshold (in grams) was calculated using the Dixon up–down method (Chaplan et al. 1994). Each hind paw was tested three times at 5-min intervals, and the mean threshold value was used for analysis. Motor coordination: Motor performance was evaluated using a rotarod apparatus (Panlab, Harvard Apparatus, Spain). Rats were trained for two consecutive days, with three sessions per day separated by 3–4-h intervals. During training, the rotation speed gradually increased from 4 to 30 rpm over 180 s. On the testing day, each rat underwent three trials separated by 5-min intervals, and the latency to fall was recorded. The average of the three trials was used for statistical analysis. Hot plate was used to test thermal hyperalgesia by using a hot plate analgesia meter (BIO-CHP, Bioseb, France) (Xu et al. 2013; Yu et al. 2008). Each rats were placed on a hot plate with the plate set to a temperature of 52 ± 0.5 °C. Response latency either to a hind-paw-lick or jump was recorded. A 20 s cut-off time was imposed to avoid tissue unnecessary damage. Western blot analysis After deep anesthesia, the L5–L6 DRGs were carefully dissected and homogenized in lysis buffer containing 25 mM Tris–HCl, 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, and 0.1% SDS (pH 7.6) supplemented with protease and phosphatase inhibitors (Halt, Thermo Fisher Scientific, Waltham, MA, USA). The lysates were gently agitated on ice for 1 h and then centrifuged at 14,000 rpm for 20 min at 4 °C. The supernatants were collected, and protein concentrations were determined using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA). Equal amounts of protein were mixed with 4 × sample buffer (250 mM Tris–HCl, pH 6.8, 8% SDS, 0.04% bromophenol blue, and 40% glycerol), heated at 95 °C for 5 min, and subjected to SDS–PAGE (Bio-Rad, Hercules, CA, USA). The separated proteins were then transferred onto PVDF membranes (Millipore, Billerica, MA, USA). Membranes were blocked for 1 h at room temperature with 5% nonfat milk or 5% BSA in TBS-T (25 mM Tris–HCl, 150 mM NaCl, 0.1% Tween-20) and subsequently incubated overnight at 4 °C with the following primary antibodies: CtBP1 (rabbit, 1:1000, GeneTex, Irvine, CA, USA; GTX111246), CtBP2 (rabbit, 1:1000, GeneTex; GTX66034), LSD1 (rabbit, 1:2000, GeneTex; GTX54328), ErbB2 (mouse, 1:200, Santa Cruz Biotechnology, Santa Cruz, CA, USA; sc-7301), and GAPDH (mouse, 1:5000, GeneTex; GTX100118). After three washes with TBS-T, membranes were incubated for 30 min at room temperature with peroxidase-conjugated secondary antibodies: goat anti-rabbit IgG (1:5000, GeneTex), goat anti-mouse IgG (1:5000, GeneTex), or goat anti-rabbit IgG (1:20,000, Jackson ImmunoResearch, West Grove, PA, USA). Immunoreactive bands were visualized using an enhanced chemiluminescence (ECL Plus) detection kit (Millipore) and imaged using the UVP ChemStudio PLUS Touch system (Analytik Jena, Germany). Densitometric quantification was performed using the corresponding analysis software. Band intensities were quantified using ImageJ software. For each sample, the signal intensities were normalized to the corresponding GAPDH band intensity to correct for loading variability. The normalized values were then expressed as fold change relative to the mean of the vehicle control group. Immunofluorescence analysis The rats were deeply anesthetized and intracardially perfused with phosphate-buffered saline (PBS) followed by 4% paraformaldehyde in PBS (pH 7.4). The L5–L6 DRGs were harvested, post-fixed in 4% paraformaldehyde at 4 °C for 3 h, and cryoprotected overnight in 30% sucrose solution. DRG Sects. (16 µm thick) were cut using a cryostat and processed for immunofluorescence staining. The sections were preincubated with 5% bovine serum albumin (BSA) and 0.01% Triton X-100 in PBS to block nonspecific binding. Subsequently, they were incubated overnight at 4 °C with the primary antibodies. After three rinses with PBS, the sections were incubated for 1 h at 37 °C with Alexa Fluor 488 (1:1500; Invitrogen, Grand Island, NY, USA) or Alexa Fluor 594 (1:1500; Invitrogen, Grand Island, NY, USA). The sections were then rinsed in PBS, mounted with coverslips, and examined under a fluorescence microscope (LEICA DM2500, Germany) equipped with a CCD camera (X-plorer; Diagnostic Instruments, Inc., USA). For multiplex immunofluorescence staining, antibodies were fluorescently labeled using the Mix-n-Stain antibody labeling kit (Biotium, Hayward, CA, USA) according to the manufacturer’s instructions. For size distribution analysis of CtBP1-positive cells within the DRG (Sharma et al. 2024; Tao et al. 2004; Xu et al. 2010), three sections were randomly selected from the first set of sections for each animal. All labeled and unlabeled cells with nuclei were counted. Cell profiles were outlined, and the cross-sectional area of each cell was calculated using SPOT 5.6 Image Capture Software (Diagnostic Instruments, USA). Primary antibodies used were as follows: rabbit anti-CtBP1 antibody (1:400; Genetex, Irvine, USA; GTX111246), mouse anti-neuronal nuclear antigen (NeuN, a neuronal marker, 1:500; Millipore, Burlington, MA, USA; 3061189), mouse anti-glial fibrillary acidic protein (GFAP, an astrocyte marker, 1:1000; Genetex, Irvine, USA; GTX34757), mouse anti-iba-1 (a macrophage marker, 1:500, Santa Cruz Biotechnology, Dallas, USA; SC-32725), mouse anti-NF-200 (a marker for A-fibre afferents and medium-large myelinated neurons, 1:1000; Sigma-Aldrich, Millipore, Burlington, MA, USA; N0142), rabbit anti-NF-200 (a marker for A-fibre afferents and medium-large myelinated neurons, 1:1000; Genetex, Irvine, USA; GTX110065), IB4 (a marker for small non-peptidergic C-nociceptors, 1:400; Sigma-Aldrich, Millipore, Burlington, MA, USA; L2895), mouse anti-CGRP (a marker for small peptidergic C-nociceptors, 1:1000; Sigma-Aldrich, Millipore, Burlington, MA, USA; C7113), rabbit anti-CGRP (a marker for small peptidergic C-nociceptors, 1:1000; Genetex, Irvine, USA; GTX134005), rabbit anti-LSD1 (1:200, Genetex, Irvine, USA; GTX54328), mouse anti-ErbB2 (1:200, Santa Cruz Biotechnology, Santa Cruz, CA, USA; sc-7301) or anti-VGLUT2 (a glutamatergic marker, mouse, 1:500, Genetex, Irvine, USA; GTX03657). Chromatin immunoprecipitation-qPCR DRGs (L5–L6) were chopped into small pieces and fixed in 1% formaldehyde to crosslink proteins with DNA. After washing, the tissues were lysed, and chromatin was broken down into fragments by sonication (200–1000 bp). A portion of each lysate was set aside as input. Chromatin was then immunoprecipitated using anti-CtBP1, anti-LSD1, anti-H3K9me2, or control IgG, followed by the addition of protein G magnetic beads. The beads were washed, and the complexes were eluted and treated with proteinase K at 62 °C, followed by reversing the crosslinks at 95 °C. DNA quantification was performed using SYBR Green-based qPCR on a QuantStudio 3 system. ErbB2_2F:GAGATCCCCAACCCCATGTG. ErbB2_2R:CCGCCACATCCTCAGGTAAA. Quantitative PCR Briefly, the dissected DRG (L5-6) was quickly removed, completely submerged in a sufficient volume of RNAlater solution (AM7021; Ambion, Rockford, IL) overnight at 4 °C to enable thorough penetration of the tissue and then maintained at 80 °C. Total RNA was isolated under RNase-free conditions using RNA isolation kits (74,106; Qiagen, Valencia, CA). Reverse transcription was performed using complementary DNA reverse transcription kits (205,311; Qiagen, Valencia, CA) Real-time PCR was carried out on a QuantStudio 3 or ABI 7500 system. Each 20 µL reaction contained 2X Master Mix and probe, with cycling conditions set at 95 °C for 20 s, followed by 40 cycles of 95 °C for 1 s and 60 °C for 20 s. RQ refers to relative quantification (RQ), which represents gene expression levels calculated by the 2 −△△CT method and normalized to GAPDH. ErbB2 cDNA_1F:GAGTCCATGCACAACCCTGA. ErbB2 cDNA_1R:TGCTGCATTTCTCACAACGC. Small-interfering RNA The siRNAs utilized in this research were sourced from Dharmacon. The sequence for CtBP1 siRNA is 5’-GCCACAUUCUGAACCUAUA-3’. These specific siRNAs were delivered intrathecally once a day for four consecutive days. Pharmacological interventions Intrathecal injections were performed via lumbar puncture at the L5–L6 intervertebral space using a 30G needle under brief isoflurane anesthesia (2–5%). AG825 (an ErbB2 inhibitor; 1, 30, and 100 nM; 10 μL; Tocris Bioscience, Bristol, UK), NSC95397 (a CtBP1 interaction blocker; 1, 10, and 30 nM; 10 μL; MedChemExpress, USA), and GSK-LSD1 (an LSD1 inhibitor; 1, 10, and 30 nM; 10 μL; Tocris Bioscience, Bristol, UK) were administered twice daily on days 11 to 13 following paclitaxel (PTX) treatment. All drugs were freshly prepared in 0.1% DMSO diluted in sterile saline, and the vehicle control group received 10 μL of 0.1% DMSO in saline according to the same injection schedule. Injections were performed slowly (over approximately 1 min) using a Hamilton microsyringe, and a tail-flick reflex was used to confirm correct intrathecal delivery. At 3 h after the last injection, behavioral studies and biochemistry analyses were conducted. Intrathecal injection procedure Intrathecal injections were performed via direct lumbar puncture at the L5–L6 intervertebral space using a 30G needle attached to a Hamilton microsyringe under brief isoflurane anesthesia (2–5%). A total volume of 10 μL was administered slowly over approximately 1 min, and a tail-flick response was used to confirm correct placement into the subarachnoid space. This procedure ensures drug delivery into the lumbar CSF, from which compounds can reach the spinal cord and adjacent DRGs through CSF diffusion. The technique followed established intrathecal injection protocols described previously (De la Calle and Paino 2002; Huang et al. 2020; Mestre et al. 1994; Yaksh and Rudy 1976). Statistical analysis Statistical Analysis was conducted using GraphPad Prism 9.0, with data expressed as mean ± SEM. Statistical comparison was performed by one-way ANOVA (using Bonferroni correction or Tukey’s tests for post hoc analysis), two-way ANOVA (using Bonferroni correction or Tukey’s tests for post hoc analysis), or unpaired Student’s t test where appropriate. Significance was set at p < 0.05. Results CtBP1 expression is increased in DRG neurons following Paclitaxel treatment To investigate the role of CtBPs in DRGs during PTX-induced neuropathic pain, male Sprague–Dawley rats received intraperitoneal PTX injections (2 mg/kg on days 0, 2, 4, and 6; total cumulative dose of 8 mg/kg; Fig. 1A). Both mechanical allodynia and thermal hyperalgesia appeared on day 7, peaked at day 14, and persisted for at least three weeks (Fig. 1B). Western blot analysis revealed a time-dependent increase in CtBP1—but not CtBP2—expression in the DRGs on days 7, 14, and 21, with a maximal rise at day 14 (Fig. 1C). Importantly, this increase was observed in both male and female rats, with no differences based on gender (Fig. 1D). Immunohistochemical analysis on day 14 confirmed increased CtBP1 immunoreactivity in DRG neurons of PTX-treated rats (Fig. 1E). PTX-enhanced CtBP1 was primarily found in NeuN (a marker for neuronal)-, NF200 (a marker for A-fibre afferents and medium-large myelinated neurons)-, IB4 (a marker for small non-peptidergic C-nociceptors)-, and CGRP(a marker for small peptidergic C-nociceptors)-positive neurons (Fig. 1F and Supplementary file 1). There was also limited co-localization in Iba-1 (a marker for macrophage)- and GFAP (a marker for satellite cell)-positive glial cells, suggesting that while CtBP1 expression is mainly neuronal, it is not exclusively so. Moreover, CtBP1 was co-expressed with VGLUT2 (a excitatory glutamatergic neurons maker) in the DRGs of PTX-treated rats (Fig. 1G), indicating its predominant localization in excitatory glutamatergic neurons within the DRG. These results suggest that neuronal CtBP1 expression is significantly upregulated in response to PTX, coinciding with the onset of mechanical allodynia and thermal hyperalgesia. PTX upregulates CtBP1 expression in the DRG neurons accompanied by neuropathic pain. Diagram of the timeline of this experiment. Paclitaxel, PTX. Vehicle, VEH.. The paw mechanical threshold and withdrawal latency in response to intraperitoneally injection of PTX or VEH in rats. Each group: 6 rats. Mechanical threshold, Two-way ANOVA, group, F(1, 50) = 40.18, p < 0.0001; time, F(4, 50) = 5.740, p = 0.0007; interaction, F(4, 50) = 6.212, p = 0.0004. Mechanical latency, Two-way ANOVA, group, F(1, 50) = 114.2, p < 0.0001; time, F(4, 50) = 11.30, p < 0.0001; interaction, F(4, 50) = 10.58, p < 0.0001. **p < 0.01 vs. VEH.. Representative Western blot and statistical analyses (normalized to GAPDH) of CtBP1 and CtBP2 protein levels in the DRGs at day −1, 3, 7, 14 and 21 after PTX or VEH injection in rats. IB, Immunoblotting. Each group: 5 rats. CtBP1, Two-way ANOVA, group, F(1, 40) = 108.2, p < 0.0001; time, F(4, 40) = 15.43, p < 0.0001; interaction, F(4, 40) = 19.36, p < 0.0001. CtBP2, Two-way ANOVA, group, F(1, 40) = 1.674, p = 0.2032; time, F(4, 40) = 0.4706, p = 0.7569; interaction, F(4, 40) = 0.1216, p = 0.9740. **p < 0.01 vs. VEH.. Representative Western blot and statistical analyses (normalized to GAPDH) of CtBP1 protein levels in the DRGs in the both gender at days −1 and 14 after PTX in rats. Each group: 5 rats. Unpaired Student’s t test. **p < 0.01 vs. male day −1. ##p < 0.05 vs. female day −1.. (F)Immunofluorescence images showing CtBP1-positive immunoreactivity (red) in the DRGs of rats treated with PTX 14D or VEH 14D. Double-labeled (yellow) images showing CtBP1 (red)- and NeuN (neuronal marker, green)-, Iba-1 (macrophages marker, green)-, GFAP (satellite cells marker, green)-, NF-200 (A-fiber afferents and medium-large myelinated neurons maker, green)-, CGRP (small peptidergic C-nociceptors maker, green)-, or IB4 (small nonpeptidergic C-nociceptors marker, green)-positive immunoreactivity. Unpaired Student’s t test. Scale bar = 100 μm. Thickness = 16 μm. All slices were harvested at day 14 after injection. Double-labeled (yellow) images showing CtBP1 (red)- and VGLUT2 a excitatory glutamatergic neurons maker, green)-positive immunoreactivity. Scale bar = 100 μm. Thickness = 16 μm. All slices were harvested at day 14 after injection CtBP1 knockdown alleviates PTX-induced mechanical allodynia To further explore the role of CtBP1 in PTX-induced pain, we performed a targeted knockdown by intrathecal delivery of CtBP1-specific siRNA (Fig. 2A). Western blot analysis demonstrated a significant reduction in CtBP1 expression in the DRGs of naïve rats following intrathecal CtBP1 siRNA administration (5 μg, 10 μL, administered once daily for four consecutive days; Fig. 2B). Compared with control siRNA administration (5 μg, 10 μL, administered once daily for four consecutive days), CtBP1 siRNA treatment also resulted in a marked decrease in CtBP1 protein levels in the DRGs. These results confirmed the efficacy of the siRNA treatment protocol. To determine whether intrathecal CtBP1 siRNA affected motor coordination or balance, the rotarod test was conducted. No significant differences in motor performance were observed between naïve, Control siRNA- and CtBP1 siRNA-treated rats (5 μg, 10 μL, administered once daily for four consecutive days; Fig. 2C), indicating that the siRNA procedure and CtBP1 silencing itself did not cause motor impairment. In PTX-treated rats, CtBP1 siRNA (5 μg, 10 μL, administered once daily for four consecutive days) significantly attenuated established mechanical allodynia and thermal hyperalgesia, while simultaneously reversing the PTX-induced upregulation of CtBP1 protein in the DRGs (Fig. 2D–F). These findings indicate that CtBP1 mediates the persistence of PTX-induced neuropathic pain. Focal knockdown of CtBP1 expression in the DRGs relieves PTX-induced neuropathic pain. Diagram of the timeline of this experiment. Paclitaxel, PTX. Vehicle, VEH. Small interfering RNA, siRNA.. Representative Western blot and statistical analyses (normalized to GAPDH) of CtBP1 abundances in the DRGs of Naïve rats in response to CtBP1 siRNA (5 μg, 10 μL, once daily for 4 days) or Control siRNA (5 μg, 10 μL, once daily for 4 days) intrathecal injection. Each group: 5 rats. One-way ANOVA, F(2, 12) = 15.79, p = 0.0004. **p < 0.01 vs. Naïve.. Rotarod performance of naïve rats (Naïve) and naïve rats treated with Control siRNA (5 μg, 10 μL, once daily for 4 days) or CtBP1 siRNA (5 μg, 10 μL, once daily for 4 days). Each group: 6 rats. Two-way ANOVA, group, F(2, 15) = 0.8773, p = 0.4362; time, F(4, 60) = 2.178, p = 0.0823; interaction, F(8, 60) = 1.019, p = 0.4317.. Paw withdrawal thresholds of VEH/PTX-treated rats following intrathecal administration of CtBP1 siRNA (CtBP1 siRNA; 5 μg, 10 μL, once daily for 4 days) or Control siRNA (Control siRNA; 5 μg, 10 μL, once daily for 4 days), given once daily on days 10–13 after PTX treatment. Each group: 5–6 rats. VEH, Two-way ANOVA, group, F(2, 15) = 1.251, p = 0.3144; time, F(4, 60) = 0.7215, p = 0.5825; interaction, F(8, 120) = 0.3439, p = 0.9433. PTX, Two-way ANOVA, group, F(2, 15) = 13.30, p = 0.0005; time, F(2.906, 43.59) = 107.8, p < 0.0001; interaction, F(5.812, 43.59) = 6.498, p < 0.0001. **p < 0.01 vs. PTX.. Paw withdrawal threshold and withdrawal latency of PTX-treated rats following intrathecal administration of CtBP1 siRNA (CtBP1 siRNA; 5 μg, 10 μL, once daily for 4 days) or Control siRNA (Control siRNA; 5 μg, 10 μL, once daily for 4 days), given once daily on days 10–13 after PTX treatment. Each group: 5 rats. Withdrawal thresholds, Two-way ANOVA, group, F(2, 9) = 43.45, p < 0.0001; time, F(3.139, 28.25) = 87.50, p < 0.0001; interaction, F(6.279, 28.25) = 7.063, p < 0.0001. Withdrawal latency, Two-way ANOVA, group, F(2, 9) = 20.71, p = 0.0004; time, F(4.033, 36.30) = 50.73, p < 0.0001; interaction, F(8.066, 36.30) = 4.012, p = 0.0017. *p < 0.05, **p < 0.01 vs. PTX.. Representative Western blot and statistical analyses (normalized to GAPDH) of CtBP1 abundances in the DRGs at day 14 of PTX rats in response to CtBP1 siRNA (PTX 14D + CtBP1 siRNA, 5 μg, 10 μL, daily at day 10–13 after PTX treatment) or Control siRNA (PTX 14D + Control siRNA, 5 μg, 10 μL, daily at day 10–13 after PTX treatment) intrathecal injection. Each group: 5 rats. One-way ANOVA, F(3, 16) = 12.64, p = 0.0002. **p < 0.01 vs. VEH. ##p < 0.01 vs. PTX CtBP1 is required for ErbB2 transcriptional activation in PTX-treated DRGs Interestingly, while CtBPs are known as corepressors, new evidence indicates they also have transactivation roles in gene regulation (Ray et al. 2014). Notably, the expression of LncRNA CtBP1-AS has been positively correlated with ErbB2 levels (Cui and Geng 2019). Overexpression of ErbB2 has been associated with both sensitivity and resistance to chemotherapy (Baselga et al. 1997). Our recent study revealed that the CtBP1-dependent pathway in the DRGs is involved in neuropathic pain caused by spinal nerve ligation (Lai et al. 2024). Therefore, we explored whether the upregulation of CtBP1 by PTX is linked to the transcriptional activation of ErbB2 in DRG neurons. Figure 3A outlines the timing of the experimental procedures used, AG825 is a selective inhibitor of ErbB2. Intrathecal delivery of AG825 (100 nM, 10 μl, administered twice daily for three consecutive days) suppressed the upregulation of ErbB2 in the DRGs on the 14th day (Fig. 3B). Administering AG825 intrathecally led to a dose-dependent reversal of mechanical allodynia (1, 30, and 100 nM; 10 μl; administered twice daily for three consecutive days; Fig. 3C, D and Supplementary file 2). Additionally, administering AG825 (100 nM, 10 μl; administered twice daily for three consecutive days) intrathecally also led to a reversal of thermal hyperalgesia (Fig. 3E). Similarly, knocking down CtBP1 with siRNA (5 μg, 10 μL, administered once daily for four consecutive days) reduced both mRNA and protein levels of ErbB2 in DRGs on day 14 of PTX-treated rats (Fig. 3F, G). Chromatin immunoprecipitation (ChIP)-qPCR assays showed increased CtBP1 binding at the ErbB2 promoter in PTX-treated rats, which was reversed by CtBP1 siRNA (5 μg, 10 μL, administered once daily for four consecutive days; Fig. 3H). These findings indicate that CtBP1 contributes to neuropathic pain through the transcriptional activation of ErbB2 in the DRGs. PTX induces CtBP1 to increase ErbB2 expression in the DRGs. Diagram of the timeline of this experiment. Paclitaxel, PTX. Vehicle, VEH. DMSO, vehicle. AG 825, a selective inhibitor of ErbB2.. Representative Western blot and statistical analyses (normalized to GAPDH) of ErbB2 abundances in the DRGs at day 14 of PTX rats in response to intrathecal injection AG 825 (a selective ErbB2 inhibitor; 100 nM, 10 μl, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment). Each group: 5 rats. One-way ANOVA, F(3 16) = 29.35, p < 0.0001. **p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D.. (D) The paw mechanical threshold measured on day 14 in PTX or VEH rats in response to AG 825 (a selective ErbB2 inhibitor; 1, 30, 100 nM; 10 μl, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment). Each group: 6 rats. C, One-way ANOVA, F(5, 30) = 40.01, p < 0.0001. D, One-way ANOVA, F(2, 12) = 0.4319, p = 0.6590. **p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D.. The paw mechanical latency measured on day 14 in PTX rats in response to AG 825 (a selective ErbB2 inhibitor; 100 nM; 10 μl, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment). Each group: 5 rats. One-way ANOVA, F(3, 16) = 31.21, p < 0.0001. **p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D.. Representative qPCR statistical analyses of ErbB2 mRNA abundances in the DRGs at day 14 of PTX rats in response to intrathecal injection CtBP1 siRNA (PTX 14D + CtBP1 siRNA, 5 μg, 10 μL, daily at day 10–13 after PTX treatment) or Control siRNA (PTX 14D + Control siRNA, 5 μg, 10 μL, daily at day 10–13 after PTX treatment). Relative quantification (RQ) = 2 −△△CT. Each group has 5 rats. One-way ANOVA, F(3, 16) = 9.901, p = 0.0006.**p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D.. Representative Western blot and statistical analyses (normalized to GAPDH) of ErbB2 abundances in the DRGs at day 14 of PTX rats in response to intrathecal injection CtBP1 siRNA (PTX 14D + CtBP1 siRNA, 5 μg, 10 μL, daily at day 10–13 after PTX treatment) or Control siRNA (PTX 14D + Control siRNA, 5 μg, 10 μL, daily at day 10–13 after PTX treatment). Each group has 5 rats. One-way ANOVA, F(3, 16) = 12.07, p = 0.0002. **p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D. ChIP-qPCR assay of the abundance of the CtBP1-precipitated ErbB2 promoter fragments at day at day 14 of PTX rats in response to intrathecal CtBP1 siRNA (PTX 14D + CtBP1 siRNA, 5 μg, 10 μL, daily at day 10–13 after PTX treatment) or Control siRNA (PTX 14D + Control siRNA, 5 μg, 10 μL, daily at day 10–13 after PTX treatment). Each group: 5 rats. CtBP1, One-way ANOVA, F(3, 16) = 42.88, p < 0.0001. IgG, One-way ANOVA, F(3, 16) = 0.63, p = 0.6051. **p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D CtBP1–protein interactions are required for ErbB2 expression in PTX-treated DRGs Transcription factors bind to gene promoters and recruit cofactors and epigenetic regulators to either activate or repress gene expression (Wang et al. 2021). We hypothesized that the upregulation of CtBP1 by PTX might require additional transcription partners to alter ErbB2 transcription and expression in DRG neurons. Figure 4A outlines the timing of the experimental procedures used. To investigate if CtBP1 needs partner proteins for ErbB2 gene activation, we used NSC95397, a small molecule that disrupts CtBP1-protein interactions (Blevins et al. 2015; Chen et al. 2021). Administering NSC95397 intrathecally (NSC; 1, 10, 30 nM; 10 μL; administered twice daily for three consecutive days) alleviated PTX-induced mechanical allodynia in a dose-dependent manner (Fig. 4B, C and Supplementary file 3). Administering NSC95397 intrathecally (NSC; 30 nM; 10 μL; administered twice daily for three consecutive days) also alleviated PTX-induced thermal hyperalgesia (Fig. 4D). In PTX rats, intrathecal NSC (30 nM; 10 μL; administered twice daily for three consecutive days) significantly reversed the PTX-induced increase in ErbB2 mRNA and protein levels, but not CtBP1, in the DRGs on day 14 (Fig. 4E, F). ChIP-qPCR revealed reduced CtBP1 binding at the ErbB2 promoter after NSC95397 treatment (30 nM; 10 μL; administered twice daily for three consecutive days; Fig. 4G). These results suggest that CtBP1-dependent transcription of ErbB2 requires interactions with protein partners in the DRGs under neuropathic conditions. PTX-induced CtBP1 upregulation requires interaction with transcription partners to regulate ErbB2 transcription/expression in DRG neurons.. Diagram of the timeline of this experiment. Paclitaxel, PTX. Vehicle, VEH. DMSO, vehicle. NSC, NSC95397, CtBP1-protein partner interaction inhibitor.. (C) The paw mechanical threshold measured on day 14 in PTX or VEH rats in response to intrathecal injection NSC95397 (NSC; inhibit the CtBP1-protein partner interaction; 1, 10 and 30 nM; 10 μL, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment) or VEH (vehicle). Each group: 6 rats. B, One-way ANOVA, F(5, 30) = 14.97, p < 0.0001. D, One-way ANOVA, F(2, 12) = 0.1286, p = 0.8805. **p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D.. The paw mechanical latency measured on day 14 in PTX rats in response to intrathecal injection NSC95397 (NSC; inhibit the CtBP1-protein partner interaction; 30 nM; 10 μL, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment) or VEH (vehicle). Each group: 6 rats. C, One-way ANOVA, F(3, 16) = 62.20, p < 0.0001.**p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D.. Representative qPCR statistical analyses of ErbB2 mRNA abundances in the DRGs at day 14 of PTX rats in response to intrathecal injection NSC95397 (NSC; inhibit the CtBP1-protein partner interaction; 30 nM; 10 μL, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment) or or VEH (vehicle). Relative quantification (RQ) = 2 −△△CT. Each group has 5 rats. One-way ANOVA, F(3, 16) = 19.32, p < 0.0001. **p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D. Representative Western blot and statistical analyses (normalized to GAPDH) of CtBP1 and ErbB2 abundances in the DRGs at day 14 of PTX rats in response to intrathecal injection NSC95397 (NSC; inhibit the CtBP1-protein partner interaction; 30 nM; 10 μL, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment) or or VEH (vehicle). Each group has 5 rats. CtBP1, One-way ANOVA, F(3, 16) = 24.92, p < 0.0001. ErbB2, One-way ANOVA, F(3, 16) = 35.69, p < 0.0001.**p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D.. ChIP-qPCR assay of the abundance of the CtBP1-precipitated ErbB2 promoter fragments at day at day 14 of PTX rats in response to intrathecal injection NSC95397 (NSC; inhibit the CtBP1-protein partner interaction; 30 nM; 10 μL, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment) or or VEH (vehicle) Each group: 5 rats. CtBP1, One-way ANOVA, F(3, 16) = 33.00, p < 0.0001. IgG, One-way ANOVA, F(3, 16) = 0.17, p = 0.9140. **p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D CtBP1 recruits LSD1 to ErbB2 promoter to drive ErbB2 expression via H3K9me2 demethylation We investigated whether CtBP1 and its associated protein LSD1 co-occupied the ErbB2 promoter, thereby removing repressive H3K9me2 marks to drive ErbB2 transcription and expression in PTX-induced neuropathic pain. Figure 5A illustrates the experimental timeline. GSK-LSD1, a selective LSD1 inhibitor, was used in this set of experiments. Co-immunoprecipitation (IP) analysis revealed that PTX increased the formation of CtBP1–CtBP1 and CtBP1–LSD1 complexes in the DRGs (Fig. 5B). Immunofluorescence imaging further demonstrated coexpression and colocalization of CtBP1, LSD1, and ErbB2 in DRG neurons of PTX-treated rats (Fig. 5C). PTX-enhanced ErbB2 immunoreactivity was mainly detected in NF200-, IB4-, and CGRP-positive neurons (Fig. 5D). Western blot analysis confirmed that PTX induced upregulation of LSD1 expression in the DRGs (Fig. 5E). Intrathecal administration of GSK-LSD1 (1, 10, or 30 nM; 10 μL; administered twice daily for three consecutive days) significantly reversed mechanical allodynia, thermal hyperalgesia, and the upregulation of LSD1 and ErbB2 in the DRGs, without affecting CtBP1 expression (Fig. 5F–I, and Supplementary file 4). Notably, CtBP1 siRNA suppressed the PTX-induced upregulation of LSD1 in the DRGs, suggesting that CtBP1 may act upstream of LSD1 (Fig. 5J). ChIP-qPCR analyses further revealed increased LSD1 binding and decreased H3K9me2 occupancy at the ErbB2 promoter following PTX treatment (Fig. 6A–C). These effects were reversed by CtBP1 siRNA (5 μg in 10 μL; administered once daily for four consecutive days) or GSK-LSD1 (30 nM in 10 μL; administered twice daily for three consecutive days). Importantly, GSK-LSD1 did not alter CtBP1 occupancy at the ErbB2 promoter, indicating that CtBP1 recruits LSD1 to mediate chromatin remodeling (Fig. 6A–C). Collectively, these data support a model in which CtBP1 forms a transcriptional complex with LSD1 to activate ErbB2 expression through H3K9me2 demethylation, contributing to the maintenance of PTX-induced neuropathic pain (Graphical Abstract). PTX enhances CtBP1/LSD1/ErbB2 signaling in DRG neurons. Diagram of the timeline of this experiment. Paclitaxel, PTX. Vehicle, VEH. DMSO, vehicle. GSK-LSD1, a selective inhibitor of LSD1. Imuncoprecipitates with CtBP1 and LSD1 in DRGs at day 14 of PTX or VEH rats. No detectable immunoreactivity was observed in the control IgG precipitation. IB, immunoblotting. In, input control. IP, immunoprecipitation. Each group has 5 rats. Unpaired Student’s t test.. Representative immunohistochemical images and statistical analyses of CtBP1-positive (blue), LSD1-positive (red), ErbB2-positive (green), and triple-labeled (white) immunofluorescence in the DRGs at day 14 of PTX or VEH rats. Scale bar = 100 μm. Thickness = 15 μm. Each group: 5 rats. Unpaired Student’s t test. **p < 0.01 vs. VEH 14D.. Double-labeled (yellow) images showing of ErbB2-positive (red) with NF-200 (A-fiber afferents and medium-large myelinated neurons maker, green)-, CGRP (small peptidergic C-nociceptors maker, green)-, or IB4 (small nonpeptidergic C-nociceptors marker, green)-positive immunoreactivity. Scale bar = 100 μm. Thickness = 16 μm. All slices were harvested at day 14 after injection.. Representative Western blot and statistical analyses (normalized to GAPDH) of LSD1 protein levels in the DRGs at days 14 after PTX or VEH in rats. Each group: 5 rats. Unpaired Student’s t test. **p < 0.01 vs. VEH 14D.. (G) The paw mechanical threshold measured on day 14 in PTX or VEH rats in response to intrathecal injection GSK-LSD1 (a selective inhibitor of LSD1, 1, 10, 30 nM; 10 μL, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment) or VEH (vehicle). Each group: 6 rats. F, One-way ANOVA, F(5, 30) = 38.77, p < 0.0001. G, One-way ANOVA, F(2, 12) = 0.1589, p = 0.8548. G, One-way ANOVA, F(2, 12) = 0.1589, p = 0.8548. **p < 0.01 vs. VEH 14D. #p < 0.05, ##p < 0.01 vs. PTX 14D.. The paw mechanical latency measured on day 14 in PTX or VEH rats in response to intrathecal injection GSK-LSD1 (a selective inhibitor of LSD1, 1, 10, 30 nM; 10 μL, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment) or VEH (vehicle). Each group: 6 rats. One-way ANOVA, F(3, 16) = 65.39, p < 0.0001. **p < 0.01 vs. VEH 14D. #p < 0.05, ##p < 0.01 vs. PTX 14D.. Representative Western blot and statistical analyses (normalized to GAPDH) of CtBP1, LSD1 and ErbB2 abundances in the DRGs at day 14 of PTX rats in response to intrathecal injection GSK-LSD1 (a selective inhibitor of LSD1, 30 nM; 10 μL, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment) or VEH (vehicle). Each group has 5 rats. The paw mechanical threshold measured on day 14 in PTX or VEH rats in response to intrathecal injection GSK-LSD1 (a selective inhibitor of LSD1, 1, 10, 30 nM; 10 μL, administered twice daily for three consecutive days from day 11 to 13 after PTX treatment) or VEH (vehicle). Each group: 6 rats. CtBP1, One-way ANOVA, F(3, 16) = 17.17, p < 0.0001. LSD1, One-way ANOVA, F(3, 16) = 37.27, p < 0.0001. ErbB2, One-way ANOVA, F(3, 16) = 13.04, p = 0.0001.**p < 0.01 vs. VEH 14D. #p < 0.05, ##p < 0.01 vs. PTX 14D.. Representative Western blot and statistical analyses (normalized to GAPDH) of LSD1 abundances in the DRGs at day 14 of PTX rats in response to intrathecal injection CtBP1 siRNA (PTX 14D + CtBP1 siRNA, 5 μg, 10 μL, daily at day 10–13 after PTX treatment) or Control siRNA (PTX 14D + Control siRNA, 5 μg, 10 μL, daily at day 10–13 after PTX treatment). Each group has 5 rats. One-way ANOVA, F(3, 16) = 14.77, p < 0.0001. **p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D PTX-enhanced CtBP1/LSD1 complex promotes ErbB2 transcription by removing repressive H3K9me2 marks in DRG neurons. (A)(B)(C) The ChIP-qPCR assay was conducted to measure the protein levels of CtBP1, LSD1, or H3K9me2-bound ErbB2 promoter fragments on day 14 in PTX-treated rats. This was in response to a selective LSD1 inhibitor (30 nM; 10 μL, administered twice daily for three consecutive days from day 11 to 13 post-PTX treatment), VEH (vehicle), CtBP1 siRNA (PTX 14D + CtBP1 siRNA, 5 μg, 10 μL, administered daily from day 10 to 13 post-PTX treatment), or Control siRNA (PTX 14D + Control siRNA, 5 μg, 10 μL, administered daily from day 10 to 13 post-PTX treatment). Each experimental group consisted of 5 rats. CtBP1, One-way ANOVA, F(3, 16) = 15.51, p < 0.0001. LSD1, One-way ANOVA, F = 22.33, p < 0.0001. H3K9me2, One-way ANOVA, F = 8.122, p = 0.0001.**p < 0.01 vs. VEH 14D. ##p < 0.01 vs. PTX 14D Discussion Initially identified as transcriptional corepressors, CtBPs interact with various transcription factors to form repressor complexes by recruiting diverse epigenetic modifiers (Lim et al. 2024). In recent years, accumulating evidence has revealed that CtBPs, in addition to their repressive function, can also promote the transactivation of specific genes (Ray et al. 2014). In the present study, we report for what we believe is the first time—the role and underlying mechanism of CtBP1 and ErbB2 in PTX-induced neuropathic pain. Our key findings are summarized as follows. First, CtBP1 was significantly and persistently upregulated in DRG neurons following PTX treatment. Second, CtBP1 facilitated LSD1-dependent H3K9me2 demethylation, thereby activating ErbB2 transcription in the DRGs of PTX-treated rats. Third, silencing CtBP1 using siRNA alleviated PTX-induced neuropathic pain. CtBP1 knockdown not only reduced pain hypersensitivity but also reversed the expression of CtBP1, LSD1, and ErbB2 in the DRGs of PTX-treated rats. Additionally, pharmacological inhibition of LSD1 mitigated neuropathic pain and reversed LSD1 and ErbB2 expression, but did not affect CtBP1 protein levels. Furthermore, since chemotherapy-induced neuropathic pain encompasses both mechanical and thermal modalities (Chen et al. 2024b) and ErbB2 signaling pathway is related to thermal pain (Jiang et al. 2014), our findings also demonstrate that CtBP1-depend ErbB2 pathway contributes to thermal hypersensitivity in addition to mechanical allodynia. Collectively, our results suggest that the CtBP1/LSD1/H3K9me2/ErbB2 signaling cascade within the DRGs serves as a key molecular mechanism underlying PTX-induced neuropathic pain. We agree that suppressing CtBP1 prior to PTX administration would help determine whether CtBP1 activation contributes to the initiation phase of chemotherapy-induced neuropathic pain. Future studies employing preventive CtBP1 knockdown or pharmacological inhibition before PTX exposure will be essential to clarify its role in the onset of neuropathy. Although our study provides molecular and histological evidence linking CtBP1–LSD1–ErbB2 signaling to excitatory neuronal activation in DRG neurons, direct electrophysiological confirmation of altered neuronal excitability remains lacking. Paclitaxel-induced neuropathy is known to involve enhanced excitatory synaptic transmission and spontaneous activity in DRG neurons. In our study, increased co-localization of CtBP1 with VGLUT2—a well-established marker of excitatory glutamatergic neurons—suggests that CtBP1 upregulation may contribute to neuronal hyperexcitability under PTX exposure. Immunofluorescence imaging further demonstrated coexpression and colocalization of CtBP1, LSD1, and ErbB2 in DRG neurons of PTX-treated rats. Nevertheless, future studies employing electrophysiological recordings (e.g., patch-clamp) or calcium imaging will be essential to validate this hypothesis and further elucidate how CtBP1–ErbB2 signaling modulates neuronal firing patterns during the development of chemotherapy-induced neuropathic pain. Although the present study primarily focused on the DRGs as a peripheral site of PTX-induced neuropathic pain, we acknowledge that the spinal cord may also contribute to the observed behavioral and molecular changes. The DRG serves as a key interface between peripheral and central nociceptive signaling, yet pain hypersensitivity is ultimately shaped by the interplay between DRG neurons, spinal dorsal horn neurons, and glial cells (Abd-Elsayed et al. 2024; Chen et al. 2024a; Jiang et al. 2025; Liu et al. 2016; Liu et al. 2014). Our ongoing experiments indicate that CtBP1 expression is also upregulated in the spinal dorsal horn following paclitaxel treatment, suggesting that CtBP1/LSD1/ErbB2 signaling may extend beyond the DRG to modulate central pain processing. Furthermore, it remains possible that pharmacological inhibition or siRNA-mediated knockdown of CtBP1 in the intrathecal space influences both DRG and spinal mechanisms through cerebrospinal fluid diffusion. Collectively, these findings highlight the broader relevance of CtBP1-associated transcriptional pathways in the development and maintenance of chemotherapy-induced neuropathic pain. Moreover, our results indicate that CtBP1 is broadly expressed across multiple neuronal subtypes in the DRG, rather than being restricted to a specific pain-related population. We acknowledge that this non-specific distribution may limit the precise attribution of CtBP1’s function to a particular subset of sensory neurons involved in pain processing. However, the functional relevance of CtBP1 in neuropathic pain is supported by our behavioral and molecular findings showing that CtBP1 knockdown alleviated paclitaxel-induced mechanical allodynia and normalized ErbB2 expression in the DRG. These results suggest that CtBP1 may act as a general transcriptional co-regulator affecting multiple neuronal populations that contribute to the overall pain phenotype. In addition, although the majority of experiments in this study were conducted using adult male Sprague–Dawley rats to minimize hormonal variability and ensure consistency across behavioral and molecular assays, we recognize this as a limitation. Considering that paclitaxel is frequently administered for female-predominant cancers such as breast and ovarian cancer (Wiseman and Spencer 1998), sex-related biological differences may influence the molecular mechanisms underlying chemotherapy-induced neuropathic pain. Indeed, studies have reported pronounced sexual dimorphism in pain signaling and immune–neuronal interactions within the DRG and spinal cord (Barcelon et al. 2023; Chernov et al. 2020). Therefore, future investigations using female models will be essential to determine whether the CtBP1–LSD1–ErbB2 signaling axis exhibits sex-dependent regulation or contributes differently to paclitaxel-induced neuropathic pain in females. The upstream mechanisms responsible for CtBP1 upregulation after paclitaxel exposure remain to be fully elucidated. Given the established role of CtBP1 as a redox-sensitive transcriptional coregulator (Deng et al. 2010), it is plausible that paclitaxel-induced oxidative stress and neuroinflammatory signaling contribute to its activation. Paclitaxel has been shown to generate reactive oxygen species (ROS) and increase proinflammatory cytokines (IL-1β, IL-6, and TNF-α) in peripheral sensory neurons and glial cells, leading to neuronal hyperexcitability and pain hypersensitivity (Fidanboylu et al. 2011; Zajaczkowska et al. 2019). Therefore, paclitaxel-induced redox imbalance and cytokine signaling may serve as upstream triggers of CtBP1 activation in DRG neurons. Future studies should clarify how oxidative and inflammatory stress converge on CtBP1 to initiate epigenetic reprogramming in chemotherapy-induced neuropathic pain. While this study identified ErbB2 as a key downstream effector of CtBP1 signaling, CtBP1 is known to regulate a broad transcriptional network that includes genes linked to inflammation, oxidative stress, and neuronal excitability. ErbB2 activation can trigger downstream PI3K/Akt, MAPK/ERK, and JAK/STAT signaling cascades, promoting neuronal sensitization and hyperexcitability (Mishra et al. 2018; Wee and Wang 2017). Thus, CtBP1-driven ErbB2 expression may amplify these pathways, sustaining neuropathic pain. Importantly, several clinically approved ErbB2 inhibitors (e.g., trastuzumab, lapatinib, neratinib, and pertuzumab) could be repurposed to mitigate paclitaxel-induced neuropathy (Graff et al. 2023). However, systemic ErbB2 blockade is associated with risks of off-target toxicity in tissues such as the heart, and the pharmacokinetic properties required to achieve sufficient DRG penetration remain unclear. Although these agents were not tested in our study, the observation that intrathecal AG825 alleviated PTX-induced allodynia in a dose-dependent manner supports the notion that ErbB2 signaling is functionally involved in neuropathic pain and may represent a viable therapeutic target. Our previous work demonstrated that CtBP1 functions as a transcriptional co-repressor in nerve injury models, where it epigenetically silences the μ-opioid receptor gene in DRGs (Lai et al. 2024). In contrast, the present study reveals that CtBP1 acts as a co-activator under paclitaxel-induced neurpathic conditions by facilitating ErbB2 transcription through cooperation with LSD1. This functional switch likely reflects the context-dependent plasticity of CtBP1, which depends on its interacting partners and local chromatin environment (Blevins et al. 2017). Elevated NADH levels and NF-κB activation triggered by PTX could promote CtBP1 dimerization and recruitment of LSD1-containing complexes (Li et al. 2018; Wang et al. 2012), thereby favoring transcriptional activation over repression in DRG neurons. Notably, Wei et al. (2022) reported no significant change in LSD1 expression in DRGs following paclitaxel treatment, which differs from our findings (Wei et al. 2022). This discrepancy may stem from differences in experimental design, time points, or tissue processing. In our study, LSD1 upregulation became evident during the sustained phase of mechanical allodynia (day 14), suggesting a delayed transcriptional response. Moreover, LSD1 activity can be modulated post-translationally rather than solely at the mRNA level. Although LSD1 is classically characterized as a co-repressor through H3K4 demethylation, accumulating evidence indicates that its function is context-dependent. When complexed with CtBP1, LSD1 can act as a co-activator by removing repressive H3K9me2 marks (Ray et al. 2014). Consistent with this mechanism, our data demonstrate that CtBP1 and LSD1 co-occupy the ErbB2 promoter, and their inhibition restores H3K9me2 levels, supporting that LSD1 functions as a co-activator to promote ErbB2 transcription in DRG neurons after paclitaxel exposure. LSD1, identified as the initial histone demethylase, plays a role in removing methyl groups from histone H3 at lysines 4 (H3K4me1/2) and 9 (H3K9me1/2) (Nair et al. 2010; Shi et al. 2004). The activation of LSD1 promotes invasion and metastasis in breast cancer by demethylating the H3K4me2 and H3K9me2 marks (Liu et al. 2020). CTBP1/2 and LSD1 function as co-activators for SREBP genes, with the CTBP-LSD1 complex's activity being modulated by ZBTB18. ZBTB18's binding to SREBP gene promoters is associated with a decrease in LSD1 demethylase activity, particularly on H3K4me2 and H3K9me2 marks (Ferrarese et al. 2023). Interestingly, alterations in the LSD1/H3K9me2 signaling pathway have been observed in ErbB2-overexpressing MCF-10A breast cells, except for H3K4me2 (Lin et al. 2021). Consequently, this study focused on H3K9me2 while investigating the CtBP/LSD1-dependent ErbB2 pathway in neuropathic pain. After nerve injury, there was a notable increase in two activating histone marks (H3K4me2 and H3K9ac) and a reduction in two repressive histone marks (H3K9me2 and H3K27me3) near the promoter region of Panx1 in DRG neurons (Zhang et al. 2015). Although H3K4me2 has been linked to neuropathic pain, it might not function through the CtBP/LSD1-ErbB2 pathway examined in this study. Further research is required to clarify its exact role in neuropathic pain. Previous investigations have suggested that paclitaxel treatment does not significantly modify ErbB2 expression levels in cancer cells, indicating that ErbB2 is not dynamically regulated by the drug in these contexts (Hayes et al. 2007). Contrary to these findings, our recent study identified a notable increase in ErbB2 expression in DRG neurons following paclitaxel administration. This inconsistency may be attributed to differences in paclitaxel dosing schedules and treatment durations. Earlier research that reported stable ErbB2 levels generally focused on cancer cells subjected to short-term or acute paclitaxel exposure, without the development of drug resistance (Haghnavaz et al. 2018; Schmidt et al. 2003). On the other hand, some studies have indicated that prolonged paclitaxel treatment can lead to drug-resistant phenotypes with heightened ErbB2 expression (Gupta et al. 2019)). Thus, the timing and duration of paclitaxel exposure appear to be pivotal factors influencing ErbB2 expression levels. Importantly, our study concentrated on non-cancerous neural tissue, specifically the DRGs, whereas most prior research was centered on ErbB2-positive breast cancer cells. Furthermore, clinical ErbB2-positive breast cancer seldom involves direct invasion or metastasis to DRG neurons. Therefore, our finding that paclitaxel induces ErbB2 upregulation in otherwise normal DRG neurons — potentially contributing to neuropathic pain — offers new mechanistic insights with direct clinical implications for managing chemotherapy-induced peripheral neuropathy. As epigenetic therapies increasingly emerge as promising treatment options for cancer and other diseases, targeting histone lysine methylation and enzyme complexes remains a significant area for therapeutic advancement (Gold and Shilatifard 2024). LSD1 and LSD2 are unique as the only FAD-dependent HKDMs. Numerous naturally occurring flavones, alkaloids, and other compounds with monoamine oxidase inhibitor (MAOI) activity have been shown or suggested to inhibit LSD1 (Fang et al. 2020). Although more selective covalent LSD1 inhibitors have encountered obstacles in clinical trials, the pursuit of developing and clinically assessing effective LSD1 inhibitors persists. The clinically validated inhibitor of lysine-specific demethylase 1 (LSD1/KDM1A) holds potential as an epigenetic therapy for ErbB2-positive breast cancer subtypes (Cuyas et al. 2020). Tolperisone, an older centrally acting muscle relaxant used for chronic pain treatment, has also been identified as a potential LSD1 inhibitor that directly demethylates target genes at H3K4me2 (Jiang et al. 2023). Research has shown that nerve injury leads to increased LSD1 protein levels in injured DRGs, contributing to neuropathic pain hypersensitivities (Wei et al. 2022). Paclitaxel has been found to elevate LSD1 expression in ovarian cancer cells, while conversely, inhibiting LSD1 enhances chemotherapy response in ovarian cancer (Chen et al. 2024c). In this study, we also discovered that paclitaxel induced CtBP1-mediated LSD1 signaling in the DRGs, which is linked to neuropathic pain. Interestingly, we observed that treatment with the LSD1 inhibitor GSK-LSD1 not only suppressed LSD1 enzymatic activity but also reduced its protein expression in DRG tissue. Although GSK-LSD1 is primarily recognized as a selective inhibitor of LSD1 catalytic function, several studies have reported that inhibition of LSD1 can disrupt the stability of its associated complex components, including LSD1 itself. Prolonged inhibition has been shown to destabilize the LSD1–CoREST complex, leading to proteasomal degradation of LSD1 and feedback suppression of its transcription (Egolf et al. 2019; Leiendecker et al. 2020). This mechanism may account for the reduction of LSD1 protein levels observed in our study following GSK-LSD1 treatment. Together, these findings suggest that both enzymatic inhibition and altered protein stability contribute to the downregulation of LSD1, further supporting the idea that CtBP1–LSD1 cooperation modulates ErbB2-dependent signaling in neuropathic pain. Therefore, we anticipate that future therapeutic developments in this field will fully consider the intricate mechanisms by which LSD1, other chromatin modifiers, and transcription factors collectively enable gene- and context-specific transcriptional regulation. Research has shown that ErbB2 is not restricted to a single neural cell type but is expressed across various populations, including neurons, astrocytes, and other glial cells. ErbB2, ErbB3, and ErbB4 transcripts and proteins are distributed throughout the adult rat brain, with distinct expression patterns in both neurons and glia (Gerecke et al. 2001; Mei and Nave 2014). Studies using human astrocytes in vitro have confirmed ErbB2 expression alongside other ErbB family members, and in vivo experiments have demonstrated that inducing ErbB2 expression in mature astrocytes drives their dedifferentiation into radial glial-like progenitors (Ghashghaei et al. 2007), emphasizing the functional importance of ErbB2 signaling in astrocyte plasticity. Moreover, ErbB2–ErbB4 signaling in hypothalamic astrocytes regulates neuroendocrine function, highlighting the physiological relevance of astroglial ErbB2 (Ma et al. 1999). In our PTX-induced neuropathic pain model, we observed a pronounced increase in ErbB2 expression in the DRG. Consistent with previous studies, ErbB2 immunoreactivity was predominantly localized to neuronal markers, indicating a major neuronal source following PTX treatment. However, we also detected low but consistent ErbB2 expression in GFAP-positive astrocytes and Iba1-positive macrophages, suggesting that glial cells may participate in ErbB2-mediated signaling under chemotherapy-induced neuroinflammatory conditions. Mechanistically, enhanced ErbB2 signaling may promote mechanical allodynia through multiple downstream pathways. ErbB2 typically heterodimerizes with ErbB3 in response to neuregulin-1 (NRG-1), a ligand known to induce pain hypersensitivity (Ma et al. 2012). Activation of ErbB2/ErbB3 can engage the PI3K/Akt and ERK/MAPK cascades, leading to increased DRG neuronal excitability, axonal sensitization, and altered expression of ion channels and pain-related genes (Jiang et al. 2014; Ledonne and Mercuri 2019). In parallel, ErbB2 signaling in glial cells can facilitate neuron–glia interactions and amplify neuroinflammatory responses through cytokine release and glial activation. Together, these mechanisms may sustain peripheral sensitization and contribute to persistent pain after chemotherapy. Overall, our findings expand the current understanding of ErbB2 dynamics in PTX-induced neuropathy by demonstrating that ErbB2 is not only upregulated in neurons but also re-expressed in glial populations. Furthermore, it is plausible that ErbB2/ErbB3 heterodimerization may also play a functional role in this process, warranting further investigation in future studies. It is essential to acknowledge that NSC95397 may exert additional biological effects beyond CtBP1 inhibition, such as influencing redox homeostasis and phosphatase activity (e.g., the Cdc25 family) or acting as an HIV-1 latency-reversing agent (Lazo et al. 2002; Nichols Doyle et al. 2024; Yang et al. 2014). These actions could potentially confound its pharmacological specificity. In our study, however, NSC95397 selectively reduced CtBP1 promoter binding without affecting CtBP1 expression, supporting its predominant role as a disruptor of CtBP1–protein interactions under our experimental conditions. Nevertheless, potential off-target effects cannot be entirely excluded. Furthermore, the translational relevance of our findings should be interpreted with caution. NSC95397 has not been clinically validated, and its pharmacokinetic and safety profiles in humans remain largely unknown. To date, this compound has been used solely as a research tool in preclinical cellular and animal studies, with no clinical trials reported. Notably, multiple clinical studies have independently reported the upregulation of CtBP1, LSD1, or ErbB2 in human tissues (Bennani-Baiti 2012; Cui and Geng 2019), where their elevated expression or co-expression has been associated with tumor progression, poor prognosis, or chemoresistance. These findings suggest that components of the CtBP1–LSD1–ErbB2 axis are clinically relevant and may be conserved across different human pathophysiological conditions, supporting the translational significance of our proposed pathway. Therefore, while our findings provide mechanistic insight into CtBP1-mediated transcriptional regulation in paclitaxel-induced neuropathic pain, future studies employing genetic knockout or knock-in models, more selective CtBP1 inhibitors, and clinically translatable compounds will be required to validate these mechanisms and evaluate their therapeutic potential in human neuropathic pain. In addition, the total PTX dose used in this study (6 mg/kg, administered as 2 mg/kg on Days 0, 2, 4, and 6) is widely employed in preclinical models of paclitaxel-induced neuropathy. Although this regimen does not directly correspond to clinical dosing in patients, it reliably induces behavioral hypersensitivity and molecular changes in the DRG, providing a practical framework for mechanistic investigation. We acknowledge that interspecies differences in pharmacokinetics may limit direct translation, and future studies will include pharmacokinetic validation to better relate these findings to clinically relevant exposures. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (PDF 1124 KB) Acknowledgements Not applicable. Authors’ contributions Performed research: H-Y P, C-Y-L, G-D C, Y-C P, C-M Y, T-B L, H-HW, J-K C, P-S Y, C-M C, D C, K-H L, S-T N, M-C H Data analysis: H-Y P, C-Y-L, G-D C, Y-C P, C-M Y Wrote the paper: H-Y P, G-D C, M-C H Designed research: M-C H. Funding This research was supported by the Ministry of Science and Technology, Taipei, Taiwan: NSTC 113–2314-B-715–003-MY3 to Dr. HY Peng, and MOST 111–2320-B-715 -003-MY3 to Dr. MC Hsieh, by the Mackay Memorial Hospital MMH-MM-111-05, MMH-MM-112-05, MMH-MM-113-03, MMH-MM-114-02, MMH-MM to Dr. HY Peng, MMH-MM-11006, MMH-MM-111-12, MMH-MM-112-13, MMH-MM-113-06, MMH-MM-114-01 Dr. MC Hsieh, and MMH-MM-112-16 to Dr. CY Lai, by as well as department of Medicine, Mackay Medical College MMC-RD-111-1B-P009 to Dr. HY Peng, MMC-RD-111-1H-P004, MMC-RD-112-1H-P001, MMC-RD-112-1B-P009, MMC-RD-113-1B-P005, MMU-RD-114-1B-P002, MMU-RD-114-1C-P003 to Dr. MC Hsieh, and MMC-RD-111-2A-P002, 24MMHIS130e and 21MMHIS466e to Dr. CY Lai, by Chung-Shan Medical University Hospital, Chung-Shan Medical University, Taichung, Taiwan CSH-2019-C-032 to Dr Chen GD as well as by department of Health, Taichung Hospital, Executive Yuan, Taichung, Taiwan 11420 and 11124 to Dr. Yeh CM. Data availability No datasets were created or analyzed during this study. Declarations Consent for publication Not applicable. Clinical trial number Not applicable. Disclosure We declare no conflicts of interest in this study. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Gin-Den Chen and Hsien-Yu Peng contributed equally to this work. References Associated Data Supplementary Materials Supplementary file1 (PDF 1124 KB) Data Availability Statement No datasets were created or analyzed during this study.