Adult Trpv1-Cre mice (JAX stock #017769) and ChR2-EYFP-report mice (JAX stock # 012569, Ai32 mice) (28), with C57BL/6 background (8-12 weeks, 25-30 g) of both sexes, were purchased from Jackson laboratories. These two lines were crossed to obtain TRPV1-ChR2 mice, wherein ChR2 is expressed in TRPV1-expressing nociceptors (29). Mice were maintained at the Animal Care Facilities of the Duke University School of Medicine. All animals were kept on a 12 hour light/dark cycle and provided with food and water ad libitum. 2-5 animals were housed per cage at ambient temperature. All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Duke University. All animal experiments were conducted according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Recombinant mouse IL-23 protein was purchased from Biolegend (Cat# 589004). Capsaicin was from Sigma-Aldrich (Cat# M2028). We obtained 17β-Estradiol (Estrogen) from Cayman Chemical (Cat# 10006315) and paclitaxel from Sigma-Aldrich (Cat# T7191). IL-23 injection contained 100 ng of IL-23 in 10 µL vehicle (PBS). Prior to IL-23 or vehicle injections, animals were briefly anesthetized with 4% isoflurane. Intraplantar injection was administered to the left hindpaw using a 30-gauge needle with a microsyringe. Spinal cord puncture for intrathecal injection was administered into cerebrospinal fluid in the region between L5 and L6 (30).

Two days prior to testing, animals were given an hour to habituate to the testing environment. Optogenetic stimulation was performed with a 470 nm LED (Thor Labs, M470F3) connected to a 1000 µm fiber (Prizmatix) as well as an LED driver (D2200, Thor Labs), used to adjust stimulus parameters. A 10 Hz blue light was used as previously described (28). Stimulation was performed by holding the 1000 µm fiber 1-2 mm below the plantar region of the left hindpaw for 20 seconds without making contact. Animal behavior was recorded using an iPhone 5S camera (30 frames/second). A resting period of 10 minutes was allotted between each of three trials. Naive animals of both sexes were stimulated at intensities of 2 mW/mm², 4 mW/mm², and 6 mW/mm² (three trials/intensity) to establish an intensity-dependence of blue light-induced pain. Naive animals of both sexes were stimulated with blue length at a low intensity (2 mW/mm²) prior to IL-23 injection and again one hour following IL-23 injection to study the effects of IL-23 on blue light-induced pain. The experimenter was blind to the treatment (vehicle vs. drug). Duration of pain was measured as the time in seconds animals spent lifting and/or lifting their afflicted paw (28).

Two days prior to testing, animals were given an hour to habituate to the testing environment. Mechanical threshold was assessed using Dixon’s Up-Down method (31), with calibrated von Frey filaments (0.16 g, 0.4 g, 0.6 g, 1.0 g, 2.0 g) (North Coast Medical). Filaments were applied to the plantar region of the left hindpaw, beginning with the lowest filament (0.16 g). If pain behavior was observed, an “X” was recorded and a lower filament was used (0.16 g remained at 0.16 g). If pain behavior was not observed, an “O” was recorded and a higher filament was used. Each animal was stimulated six times at each of three time points (baseline, one hour, and three hours following injection). The order of Os and Xs recorded were then converted to mechanical thresholds (in grams).

Animals (5 males and 5 females) were briefly anesthetized with 4% isoflurane, then perfused through the left ventricle PBS and 4% PFA using 30-mL syringes. Following perfusion, L4 and L5 DRGs were collected and postfixed in 4% formaldehyde at 4°C for 2 hours. DRGs were placed in a 30% sucrose solution and dehydrated overnight. DRG tissues were then embedded in OCT medium (Tissue-Tek), sectioned (14 μm) in a cryostat, and thaw-mounted onto Superfrost Plus slides (VWR). In situ hybridization was performed using the RNAscope system (Advanced Cell Diagnostics) following the manufacturer’s instructions. Probe for murine Trpv-1 (313331-C2) was applied in this study. DRG sections were dehydrated by washing in PBS (2 x 5 min) and placing in an increasing alcohol gradient, then pre-treated with hydrogen peroxide, washed, and pre-treated with Protease IV. The Trpv1 probe (mixed with RNAscope Diluent, 1:50) was added to the sections, which were heated to 40°C for 2 hours, then washed with wash buffer (2 x 2 min). The probe was amplified using Amp 1, Amp 2, and Amp 3 solution, with washing after each addition. HRP-C2 was then added to the sections. After washing with wash buffer, Fluorescein-C2 was added (CY5, 1:750), and the sections were again heated to 40°C for 30 min. Following additional washing, HRP blocker was added to the sections, which were heated to 40°C for a third time for 15 min. Slides were again washed with wash buffer, then mounted with mounting media. All images were acquired with the same settings using a Nikon fluorescence microscope under 20x magnification. We selected four non-adjacent DRG sections from each animal and included five animals per sex for data analysis. The sizes of Trpv1-positive neurons were analyzed blindly using Image J.

Immunohistochemistry of IL-23 and macrophage marker F4/80 was performed on naïve animals of both sexes. Immunohistochemistry of P-p38 was performed on animals of both sexes who had undergone intrathecal injection of vehicle or IL-23. Thirty minutes after injection, animals were anesthetized with isoflurane and perfused through the left ventricle with first PBS, then 4% formaldehyde. After perfusion, the L4-L5 DRGs were collected and postfixed in 4% formaldehyde at 4°C for 2 hours. The DRGs were then placed in a solution of 30% sucrose in PBS at 4°C and dehydrated overnight. The DRGs were mounted with optimal cutting temperature medium (Tissue-Tek), then cut with a cryostat (Leica) to a thickness of 14 μm and thaw-mounted onto Superfrost Plus slides (VWR). The sections were blocked with blocking buffer (5% donkey serum and 0.1% Triton X-100 in PBS) for 1 hour at room temperature. Afterward, primary antibodies were diluted in 1% BSA and 0.2% Triton X-100; the sections were then incubated with the primary antibodies overnight at 4°C. The primary antibodies include IL-23 monoclonal antibody (mouse, 1:500, Santa Cruz, sc-271219), F4/80 monoclonal antibody (rat, 1:100, Invitrogen, PA5-21399), and p-p38 (rabbit, 1:500, Cell Signaling, #9212). For double-staining of IL-23 and F4/80, two primary antibodies from different species were mixed. Following incubation, the sections were washed with PBS (3 x 5 min), and incubated with secondary antibodies (1:500, Jackson ImmunoResearch) for 1-2 at room temperature. The secondary antibodies include 488-conjugated anti-mouse, cy3-conjugated anti-rat, and cy3-conjugated anti-rabbit antibodies. For double staining, two secondary antibodies were also mixed. After washing with PBS again (3 x 5 min), the sections were mounted with a coverslip and the mounting medium, Prolong Gold (Life Technologies), and allowed to dry overnight at room temperature. Images were obtained using a Leica Confocal Microscope at the Duke University Microscope Center. The number of IL-23 and F4/80 double-positive cells in 4 DRG sections per animal was quantified blindly and calculated as the number of cells per mm² of tissue.

THP-1 cells (human macrophage cell line) were purchased from the Duke core facility (ATCC # TIP202). Cells were cultured in high glucose (4.5 g/L) RPMI 1640 Medium containing 10% (v/v) FBS at 37°C. THP-1 cells were treated in RPMI media containing 250 nM PMA (Sigma) for 48 hours for differentiation and then washed with PBS to eliminate non-adherent cells. The adherent cells were differentiated (mature) human macrophages. To examine IL-23 and IL-17A secretion in THP1 cells, we purchased hIL-23 ELISA kit from Biolegend (# 437607, San Diego, CA) and hIL-17A ELISA kit from Meso Scale Diagnostics (# K151RFD-1, Rockville, MD). After seeding matured THP1 cells (500,000, 400 μl), we treated these macrophages with β-estradiol (1 ng/ml), paclitaxel (1 µg/ml), and their combination for 24 hours at 37°C. We collected 200 μl of the culture media for ELISA assays. We measured secreted cytokine levels using a Bio-Rad plate reader for hIL-23 and a MESO QuickPlex SQ 120 for hIL-17A, according to the manufacturer’s protocol.

All data in this study are expressed as mean ± SEM. All sample sizes are included in figure legends. Animals were randomly assigned to each experimental group. The results of the behavioral tests were analyzed with two-way ANOVA, or one-way ANOVA, followed by Bonferroni’s post-hoc test, or by unpaired Student’s t-test. In cases when the data appeared to be non-normally distributed, a normality test was performed (D’Agostino-Pearson omnibus normality test), followed by a non-parametric Mann-Whitney test or Kruskal-Wallis test. Differences between groups were deemed statistically significant if the p-value was less than 0.05. Statistical significance was labeled as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Our recent study showed comparable expression of Trpv1 mRNA in male and female mice (Luo et al., 2021). Given the important role of TRPV1 in immune cells and nociceptor interactions, we conducted additional analysis of Trpv1 expression using a cohort of 10 male and female mice (n = 5 for each sex). We used a highly selective and sensitive RNAscope protocol (Advanced Cell Diagnostics) to assess the relative numbers of TRPV1⁺) neurons in the DRG of mice of both sexes ( Figures 1A, B ). We found that females appear to have a slightly greater number of Trpv1 ⁺ neurons than males, albeit not to a significant extent (female: 30.70 ± 3.48; male: 24.90 ± 1.55, per arbitrary unit, p = 0.1825, n = 5 mice/sex) ( Figure 1C ). We also found that the average size of Trpv1 ⁺ neurons was slightly but significantly smaller in female mice vs. male mice (female: 262.68 ± 6.28 µm²; male: 283.21 ± 10.22 µm², p < 0.05, n = 5 mice/sex) ( Figure 1D ). We then assessed the sizes of the Trpv1 ⁺ neurons in both sexes, organizing the neurons into size categories of 0-100 µm², 100-200 µm², 200-300 µm², up to 700-800 µm² ( Figure 1E ). To account for the difference in the number of Trpv1 ⁺ neurons for each sex, we expressed our quantifications as percentages, calculated as the number of Trpv1 ⁺ neurons in the size category for the specified sex over the total number of Trpv1 ⁺-positive neurons for the specified sex. Strikingly, females had a significantly higher percentage than males in the size category of 100-200 µm² (female: 31.11 ± 1.48%, male 24.57 ± 3.28%, p < 0.05, n = 5 mice/sex, Figure 1E ). However, there was no significant difference between females and males for the other size categories, although males appeared to have larger populations in the 300-400 and 400-500 µm² size ranges ( Figure 1E ). Collectively, these results suggest that Trpv1 ⁺ may be present in more small-sized nociceptive neurons in females compared to males.

Optogenetic approaches have been used to activate or suppress nociceptors for pain modulation (29, 32). However, sex-dimorphism of light-stimulated pain has not been investigated. We first tested blue light-induced pain in both sexes by stimulating TRPV1-ChR2 mice at three different intensities (2 mW/mm², 4 mW/mm², and 6 mW/mm²) and measuring the duration of the resulting pain behavior (28, 29). Figure 2A shows the experimental paradigm of light stimulation. We found that blue light-induced spontaneous pain is intensity-dependent: mice of both sexes showed a significantly higher duration of pain response with each increase in intensity (P <0.0001, Figure 2B ). We did not observe sex differences in light-induced pain at the intensities of 2 mW/mm² and 4 mW/mm² (p > 0.05, Figure 2B ). Interestingly, females showed a significantly higher duration of pain response than males at the highest intensity (6 mW/mm², p < 0.0001, Figure 2B ).

We then examined whether IL-23 could potentiate blue light-induced pain in mice of both sexes ( Figure 2C ). To see an additive effect of two treatments, we chose a low intensity of light stimulation (2 mW/mm²). We found that while durations of pain response were comparable for mice of both sexes at baseline, females demonstrated significantly higher durations of pain response at one hour following intraplantar IL-23 injection (100 ng), whereas males did not ( Figure 2D ). These findings suggest that IL-23 potentiates blue light-induced pain in only female mice.

Our recent study demonstrated that IL-23 was sufficient to induce mechanical pain in female but not male mice (27). We found that intraplantar IL-23 (100 ng) could induce lasting mechanical pain (over 5 hours) in females, but not in males, using von Frey testing. To further understand the role of IL-23 in the sex dimorphism of mechanical pain and to confirm the findings of Luo et al., we used the von-Frey Up-Down test to assess the withdrawal thresholds of mice of both sexes at baseline, one hour, and three hours after intraplantar IL-23 injection (100 ng). We found that while the withdrawal thresholds of males injected with IL-23 did not change significantly, the withdrawal thresholds of females injected with IL-23 decreased significantly one hour and three hours following injection. Females injected with IL-23 had significantly lower paw withdrawal thresholds than their vehicle counterparts at both time points (p < 0.001 at 1 h at 3 h, Figure 3A ). Our finding confirms that IL-23 does participate in the sex dimorphism of mechanical pain.

Intense and persistent activation of C-fibers, such as TRPV1 and TRPA1 agonists, is known to produce remarkable edema and neurogenic inflammation (33, 34). Our recent study shows necessity of TRPV1 for IL-23-induced mechanical pain in females (27). To this end, we measured paw thickness 3 hours after intraplantar injection of IL-23 (100 ng) or vehicle (PBS) in male and female mice. We found no difference in paw thickness in vehicle and IL-23 injected paws in both sexes ( Figure 3B ). This finding suggested that IL-23 does not induce edema, indicating different mechanisms of TRPV1 activation by capsaicin (direct mechanism) vs. IL-23 (indirect mechanism).

Capsaicin produces acute spontaneous pain, such as licking and lifting of the affected paw, following intraplantar injection in mice (35). We also examined the role of IL-23 in the sex dimorphism of spontaneous pain as induced by capsaicin. In order to see a potentiation by IL-23, we tested a low-dose of capsaicin (500 ng). Intraplantar capsaicin was given 1 h after intrathecal injection of 100 ng of IL-23 or vehicle (PBS) ( Figure 4A ). While intraplantar IL-23 targets macrophages in hind paw skin, intrathecal IL-23 could target macrophages in DRG (27). IL-23 can enhance TRPV1 activity both in DRG neuron cell bodies (mediated by intrathecal injection) and in skin nociceptor terminals (mediated by intraplantar IL-23). We have shown that either intraplantar or intrathecal IL-23 is sufficient to induce mechanical pain in female mice (27). We found that males given vehicle (PBS) and males given IL-23 showed comparable durations of pain response following capsaicin injection (p > 0.05, Figures 4B, C ), while females given IL-23 showed significantly higher durations of pain response than their vehicle counterparts (p < 0.05, Figures 4D, E ). These results suggest that IL-23 potentiates capsaicin-induced spontaneous pain in only females and not in males. Since IL-23 was given by intrathecal injection and IL-23 does not act on nociceptors directly (27), IL-23 may indirectly sensitize TRPV1 at DRG cell bodies to potentiate intraplantar capsaicin-evoked spontaneous pain.

We used immunohistochemistry to examine IL-23 expression in DRGs of male and female mice. IL-23 immunoreactivity was found in DRG sections of both sexes, and we did not see IL-23 expression in DRG neurons ( Figure 5A ). Double staining revealed a high degree of co-localization of IL-23 and F4/80, a marker for DRG macrophages (Luo et al., 2019b), in both sexes ( Figure 5A ). Further quantitative analysis revealed that females had a slightly higher number of IL-23-positive macrophages than males, albeit not to a significant extent (218.42 ± 14.04/mm² in females vs. 203.53 ± 35.68/mm² in males, p = 0.7180, Figure 5B ). These results indicate the presence of IL-23 in DRG macrophages of naïve mice, providing a cellular base for IL-23-mediated macrophage and nociceptor interaction.

Phosphorylation of p38 MAP kinase (p-p38) is a marker for nociceptor activation, as p-p38 is induced in DRG neurons by noxious stimulation, inflammation, and nerve injury and contributes critically to the pathogenesis of pain (36–39). To study the role of IL-23 in pain signaling in DRG neurons, we used immunohistochemistry to stain for p-p38 in the DRG of male and female mice ( Figure 6 ), without IL-23 treatment ( Figures 6A, B ), and with IL-23 treatment ( Figures 6C–E ). In support of previous reports (36, 37), p-p38 was primarily present in the nuclei, and the basal levels of p-p38 in DRG neurons were lower in both sexes, labeling 5-15% of DRG neurons in both sexes. Female DRG had a slightly higher basal levels compared to male DRG (p > 0.05, Figure 6F ). In order to target DRG cells bodies, we injected IL-23 (100 ng) via intrathecal (I.T.) route (36, 40, 41). Thirty minutes following IL-23 treatment, p-p38⁺ neurons increased in both sexes (p < 0.0001), labeling 20-40% of DRG neurons in both sexes ( Figure 6F ). However, female DRGs had a significantly higher percentage of p-p38-positive neurons than male DRGs (p < 0.05, Figure 6F ). After IL-23 treatment, the majority of p-p38-positive cells were small-sized DRG neurons (presumably nociceptors, Figure 6E ), indicating an activation of nociceptors by IL-23. In addition, we also observed p-p38 expression in the nuclei of non-neuronal cells in female DRG, including satellite glial cells and immune cells ( Figure 6E ). Taken together, these results suggest that IL-23 can promote pain via activation of nociceptors in a sex-dependent manner. Since p-p38 is highly co-localized with TRPV1 in DRG neurons under inflammatory conditions and further regulates TRPV1 expression and activity (14, 37), p38 activation may also contribute to IL-23-induced pain.

Our recent study showed that IL-23 does not activate nociceptors directly. Instead, IL-23 produces IL-17A in macrophages to activate nociceptors (Luo et al., 2021). To further address the question of IL-23 release and IL-23 signaling in macrophages, we conducted ELISA analysis in THP-1 cells, using a well-characterized cell line of human macrophages. THP-1 cells constitutively released IL-23, with a baseline ranging from 6 to 90 ng/ml from 500,000 cells with 400 μl medium ( Figure 7A ). We stimulated THP-1 cells with estrogen (17β-Estradiol, 1 ng/ml) and paclitaxel (PTX, 1 µg/ml), a chemotherapy drug that can induce neuropathic pain. PTX was also shown to act on peritoneal macrophages to produce inflammatory cytokines (TNF and CXCL1) (42). Application of low dose PTX or estrogen (24 h) did not produce significant increase in IL-23 secretion, but a combination of PTX and estrogen significantly increased IL-23 release in culture medium (p < 0.05, vs control or PTX alone, Figure 7A ).

We recently showed that IL-17A was required for IL-23-induced mechanical pain (Luo et al., 2021). At the baseline, THP-1 cells secreted low levels of IL-17A (3-7 ng/ml), but this secretion was significantly increased by co-application of PTX (1 μg/ml, 24 h) and estrogen (1 ng/ml, 24 h), (p < 0.05, vs control, Figure 7B ). Interestingly, stimulation with IL-23 (10 ng/mL, 24 h) alone did not increase IL-17A secretion, but co-application of IL-23 (10 ng/ml, 24 h) and estrogen (1 ng/ml, 24 h) did (p < 0.05, vs. control and estrogen, Figure 7C ). Considered together, these results suggest (1) THP-1 macrophages secrete IL-23 and IL-17A and (2) this secretion is increased by estrogen and chemotherapy. These results also suggest IL-17A may be downstream of IL-23 signaling in macrophages.