All animal procedures were approved by the Animal Ethics Committee at Nanjing Medical University. Adult male and female C57BL/6 mice (Laboratory Animal Center of Nanjing Medical University, Nanjing, China) were housed with ad libitum access to food and water. Newborn male pups were separated from their dams in the same cage by an opaque isolation board for 3 h daily from 9:00 to 12:00 h from postnatal days 2–15, then the board was removed and pups were returned to their dams (Du et al., 2019). Control pups were raised under the same conditions except that no isolation board was introduced into the cage. All pups were weaned and separated from their dams on postnatal day 21, and analyzed as described below starting at 8 weeks of age.

The numbers of animals and replicate measurements in different experiments were based on our previous experience (Ruan et al., 2010; Cao et al., 2008).

Mice that experienced maternal separation or not were anesthetized intraperitoneally using sodium pentobarbital (40 mg/kg), then injected intrathecally with one of the following: a chimera of EphB2 with Fc (catalog no. 467-B2, R&D Systems, Minnesota, United States), which binds to EphrinB2 and prevents it from stimulating endogenous EphB2; a chimera of EphrinB2 and Fc (catalog no. 496-EB, R&D Systems, Minnesota, United States), which binds and activates endogenous EphB2; or, as a negative control, recombinant Fc from human IgG1 (catalog no. 110-HG, R&D Systems, Minnesota, United States). Injections EphB2-Fc or EphrinB2-Fc(0.1-µg or 0.5-µg/5 μL, i. t., daily for three consecutive days) were made between vertebrae L5 and L6 using a 25-µL Hamilton microinjector (Ruan et al., 2010), and tail flicking was interpreted as successful puncture. Injections were delivered within 30 s, the needle was left in place for 15 s, then the needle was gently removed. All doses of the drug are based on the results of the pre-experimental studies in reference to Zhou et al. (2015a). The detailed dose and treatment time point for each drug are given in the illustration section.

Visceral sensitivity was quantified by scoring the abdominal withdrawal reflex following colorectal distention (Li et al., 2023). Mice were lightly anesthetized with 1%–1.5% isoflurane, then a rubber balloon was carefully inserted into the colorectum 3 cm from the anus and secured to the base of the tail using hypoallergenic silk tape. Animals were placed individually into a clear box (20 × 10 × 18 cm), allowed to habituate for 10–15 min, then the balloon was inflated sequentially for 20 s at pressures of 20, 40 and 60 mmHg. The balloon was deflated for 2–3 min between inflations. The abdominal withdrawal reflex was scored from 0 to 4 as described (Chen et al., 2022) by investigators blinded to mouse grouping. Scores were averaged from triplicate measurements on each animal.

In immunohistochemistry, immunofluorescence staining and Western blotting experiments to compare activation or expression of key proteins between mice that experienced maternal separation or not (see below), colorectal distension stimulation was performed at 60 mmHg every 3 min during 15 min. We chose these conditions to avoid damaging the colorectal tissue, given that the balloon had to be inflated at 60 mmHg in order to induce nociceptive behaviors.

Mice were anesthetized under isoflurane before transcardial perfusion and injected with saline and 4% paraformaldehyde, then the spinal cord was removed and sliced into sections 20 µm thick using a cryostat (CM1800, Leica, Germany). Sections were washed in phosphate-buffered saline (PBS), blocked in goat serum for 30 min, then incubated with primary antibodies against the following proteins: EphB2 (1:200; catalog no. MABN726, Merck & Millipore, United States), EphrinB2 (1:100; catalog no. HPA008999, Merck & Millipore, United States), phospho-ERK (1:200; catalog no. sc-136521, Santa Cruz, Dallas, United States), phospho-p38 (1:200; catalog no. sc-166182, Santa Cruz, Dallas, United States), phospho-JNK (1:200; catalog no. sc-6254, Santa Cruz, Dallas, United States), c-fos (1:500; catalog no. sc-166940, Santa Cruz, Dallas, United States), GFAP (1:200; catalog no. sc-33673, Santa Cruz, Dallas, United States), NeuN (1:1500; catalog no. ABN90, Merck & Millipore, United States), and Iba1 (1:200; catalog no. sc-32725, Santa Cruz, Dallas, United States). Subsequently, the sections were rinsed three times with PBS and incubated for 2 h with anti-rabbit Alexa Fluor 488 (1:500, Invitrogen, California, United States) or anti-mouse Alexa Fluor 555 (1:500, Invitrogen, California, United States) antibodies. After washing with PBS for 3 times, incubate with DAPI.

Immunostained sections were examined under Olympus microscope (1–71, Olympus, Japan) and images were analyzed using ImageJ software (National Institutes of Health, United States).

Under isoflurane anesthesia, mice were sacrificed, the spinal cord was removed, flash-frozen in liquid nitrogen, then thawed and lysed in RIPA lysis buffer. Equal amounts of total protein (30 µg) were fractionated by SDS-PAGE and transferred onto polyvinylidene difluoride membranes, which were immersed in blocking buffer (Beyotime, Shanghai, China), then incubated with primary antibodies against the following proteins: EphB2 (1:500; catalog no. MABN726, Merck & Millipore, United States), EphrinB2 (1:500; catalog no. HPA008999, Merck & Millipore, United States), phospho-ERK (1:500; catalog no. sc-136521, Santa Cruz, Dallas, United States), ERK (1:500; catalog no. sc-514302, Santa Cruz, Dallas, United States), phospho-p38 (1:500; catalog no. sc-166182, Santa Cruz, Dallas, United States), p38 (1:500; catalog no. sc-7972, Santa Cruz, Dallas, United States), phospho-JNK (1:500; catalog no. sc-6254, Santa Cruz, Dallas, United States), JNK (1:500; catalog no. sc-137019, Santa Cruz, Dallas, United States), GFAP (1:200; catalog no. sc-33673, Santa Cruz, Dallas, United States), NeuN (1:2000; catalog no. ABN90, Merck & Millipore, United States), Iba1 (1:500; catalog no. sc-32725, Santa Cruz, Dallas, United States), NR2B (1:1000; catalog no. 42125, Cell Signaling Technology, Massachusetts, United States), phospho-NR2B (TYR1472) (1:1000; catalog no.4208, Cell Signaling Technology, Massachusetts, United States), c-fos (1:1000; catalog no. sc-166940, Santa Cruz, Dallas, United States), and Actin (1:2000; catalog no. sc-8432, Santa Cruz, Dallas, United States). Membranes were washed three times with Tris-buffered saline containing 0.1% Tween-20 (TBST), then incubated for 2 h at room temperature with horseradish peroxidase-conjugated goat secondary antibodies against mouse IgG (1:5000; catalog no. AP124P, Merck & Millipore, United States) or rabbit IgG (1:5000; catalog no. AP132P, Merck & Millipore, United States). Membranes were again washed three times with TBST, images were scanned with Micro Chemi (DNR Bio-imaging systems, Jerusalem, Israel) and the optical densities of detected bands were quantified using Image software J (US National Institutes of Health, Bethesda, MD, United States).

Data were analyzed and figures prepared using Graph Pad Prism 9.0.0 (Inc., San Diego, CA). Data were reported as mean ± SEM unless stated otherwise. Pairwise differences were assessed for significance using the two-tailed unpaired Student’s t-test, while differences among three or more groups were assessed using one- or two-way analysis of variance. Differences associated with p < 0.05 were considered significant.

Mice were subjected to colorectal distension (CRD) following maternal separation (MS). The schematic diagram and time course of CRD combined with abdominal withdrawal reflex (AWR) score under pressures of 20, 40, and 60 mmHg are shown in Figure 1A . The records confirmed that separating mouse pups from their mothers for a 14-day period soon after birth led to visceral hyperalgesia several weeks later, which we detected as an increase in abdominal withdrawal reflex in response to colorectal distension, the AWR score of MS mice did not change at 20 mmHg but increased significantly at 40 and 60 mmHg (Figure 1B). The stronger reflex was associated with upregulation of c-fos (Figure 1C), which signals neuronal activation during nociception (Coggeshall, 2005), as well as with upregulation of EphB2 and EphrinB2 (Figure 1D). The time points for the detection of EphrinB2/EphB2, and Fos protein expression are set at 60 min and 90 min after CRD, respectively. In fact, levels of c-fos correlated positively with those of EphB2 and EphrinB2 (Figure 1E). These results suggest that stress due to maternal separation induces visceral hyperalgesia by activating EphrinB2/EphB2 signaling.

Visceral hyperalgesia due to maternal separation was associated with activation of three MAP kinases known to act downstream of EphB2: ERK, p38 and JNK (Figure 2A). Immunofluorescence staining of thin sections from the spine showed that EphB2-positive cells in the spinal cord mainly co-expressed with these kinases 60min after CRD (Figures 2B–D). In the control group, we observed a small amount of EphB2 or MAPKs activation, but the co-expression of these receptors were significant less than those in MS group (Figure 2E). The ratio of EphB2 + p-ERK, EphB2 + p-p38 and EphB2 + p-JNK to EphB2 was 70.02%, 61.46% and 76.05%, respectively (Figure 2F, left). While the ratio of EphB2 + p-ERK to p-ERK, EphB2 + p-p38 to p-p38 and EphB2 + p-JNK to p-JNK was 72.16%,74.36% and 71.93%, respectively (Figure 2F, right). These results suggest that the stress caused by maternal separation stimulates EphrinB2/EphB2 signaling, which in turn activates MAP kinases.

Visceral hyperalgesia due to maternal separation was associated with upregulation of EphB2 and EphrinB2 in three key cell types in the dorsal horn of the spine: neurons, which we identified based on immunostaining against NeuN; astrocytes, based on immunostaining against GFAP; and activated microglia, based on immunostaining against Iba1 in the dorsal horn at 60min (Figures 3A, B). We also observed that a small amount of EphB2 or EphrinB2 was activated in neurons, microglia or astrocytes in the control group, but the amount of co-expression was significantly lower than those in the MS group (Figures 3C, D). The proportion of EphB2 activation in neurons, microglia and astrocytes was 67.37%, 10.14% and 22.23%, respectively. While the proportion of EphrinB2 activation in neurons, microglia and astrocytes was 59.83%, 13.94% and 26.24%, respectively (Figures 3E, F). These results suggest that visceral hyperalgesia induced by MS activates the EphrinB2/EphB2 signaling pathway in neurons and is accompanied by activation of this signaling pathway in microglia and astrocytes.

All our experiments so far pointed to EphrinB2/EphB2 signaling as a driver of visceral hyperalgesia due to maternal separation. Consistent with this idea, pretreatment with repeated administration of EphB2-Fc (0.5-µg, i. t., daily for three consecutive days, see experimental protocol in Figure 4A) before CRD inhibited such signaling in mice that had experienced maternal separation mitigated their abdominal withdrawal reflex in response to colorectal distension (Figure 4B), while also downregulating c-fos (Figure 4C). In support of the reliability of our results, inhibiting EphrinB2/EphB2 signaling did not significantly alter the abdominal withdrawal reflex in mice that had not experienced maternal separation, nor did it alter c-fos expression in the absence of colorectal distension.

These therapeutic effects of inhibiting EphrinB2/EphB2 signaling were associated with reduced activation of the three MAP kinases ERK, p38 and JNK in the spine (Figure 4D), as well as lower overall numbers of neurons, astrocytes, and activated microglia (Figure 4E). Given that the NMDA receptor appears to mediate at least some nociceptive signaling by EphrinB/EphB (Liu et al., 2013), we analyzed phosphorylation of the NMDA receptor subunit NR2B in the presence or absence of EphrinB2/EphB2 inhibition. Such inhibition partially reversed the phosphorylation of NR2B induced by colorectal distension, without affecting total levels of NR2B (Figure 4F).

These results support the idea that EphrinB2/EphB2 signaling helps mediate visceral hyperalgesia due to maternal separation, at least in part by stimulating the NMDA receptor.

The above results suggested that the activation of endogenous EphB2 receptor promotes activation of glial cells and neurons, and produced viscera hyperalgesia after CRD in MS mice. NMDA receptor may contributed to the activation of glial cells and neurons in this process. A corollary of this idea is that activation of EphrinB2/EphB2 signaling should induce changes in naïve mice similar to those observed in animals that experienced maternal separation. We tested this idea by giving naïve mice exogenous EphrinB2 (0.1-µg or 0.5-µg, i. t., daily for three consecutive days, see experimental protocol in Figure 5A). As predicted, the resulting activation of EphrinB2/EphB2 signaling led to visceral hyperalgesia as measured following colorectal distension (Figure 5B), and in the spine it activated the three MAP kinases (Figure 5C); increased numbers of neurons, astrocytes and activated microglia (Figure 5D); and increased levels of phosphorylated NR2B and c-fos (Figure 5E), especially in the higher dose (EphrinB2-Fc 0.5-µg) group. These effects were stronger in animals given the higher dose of exogenous EphrinB2, indicating dose dependence.

These results strongly support a model of visceral hyperalgesia in which EphrinB2/EphB2 signaling in the spine acts via MAP kinases to alter interactions among neurons, microglia and astrocytes.