Twenty-three pregnant Wistar rats were purchased from B&K Universal Group Ltd. (Shanghai, China). All pregnant rats were housed separately at a constant temperature with light from 7:00 am to 7:00 pm and had unlimited access to food and water. The day of delivery was recorded as PND0. After giving birth, the mothers were kept together with their newborn offspring to feed and take care of them until PND21. To rule out the effects of the estrous cycle, only male pups were involved in this study (n = 157). However, 11 pups died during the MS procedure, and four rats died during the microinjection procedure. Therefore, 142 male pups were finally used for further experiments in this study (study design shown in Figure 1). All animals were treated in accordance with the guidelines for the Care and Use of Laboratory Animals, and all procedures were approved by the Animal Research Committee of the Tongji University, Shanghai, China.

A rat model of MS reported in previous research by the authors was established (Yi et al., 2017). In summary, the pups were randomly assigned to the MS and normal control (NC) groups on PND1. During PND2-PND14, the MS pups were moved gently from their mother and siblings at 10:00 am to separate small plastic boxes (19 cm × 10 cm × 14 cm) on a heating pad (37 ± 0.5°C) in another room, suffering a 3 h-maternal deprivation daily. NC pups were left undisturbed. All pups were weaned and sex-determined on PND21. Both groups of pups were randomly subdivided into post-weaning (PND21), prepubertal (PND35), and adult (PND70) groups separately, which determines the time they would be disposed of. Five to six male rats of the same subgroup were then kept in one cage until the experiments were performed.

Anxiety-like behavioral tests were performed by an observer blinded to the identity of the experimental groups. Each test was performed once on each rat. The results were measured using the EthoVisionXT8.0 Tracking System (Noldus Information Technology, Wageningen, the Netherlands). To adapt to the environment, animals were transferred into the specific testing room 24 h before the actual tests. The floor and the testing apparatus were cleaned with 75% ethanol and dried between two tests.

Visceral sensitivity was assessed by recording animals’ visceral motor response (VMR) to colorectal distension (CRD), which was described in a previous study (Yi et al., 2017; Yi et al., 2019). When behavioral tests were completed, rats were anesthetized with an intraperitoneal injection of pentobarbital sodium (50 mg/kg), then a pair of Teflon-coated silver wires were implanted in the left obliques externus abdominis 1 day before EMG recordings (study design shown in Figure 1A). The rats were caged separately for rest and fasted overnight before the assessment. On the ensuing visceral sensitivity test day, each rat was placed into a well-sized oval retainer made of Perspex (Yuyan Instruments, Shanghai, China). The EMG signal was recorded by the Amplifier (Brownlee Precision Model 440, United States) and Axon Digidata 1440A data acquisition system (Molecular Devices, United States). A pre-made latex balloon (0.5 cm diameter for rats on PND21, 1 cm diameter for rats on PND35, 2 cm diameter for rats on PND70) tied around a medical catheter (2 mm diameter) was inserted into the colorectal cavity and fixed in place with surgical tape around the tail. The baseline EMG was recorded for the first 20 s. Then gas was injected within 1 s to maintain a certain pressure for the next 20 s. CRD was performed at graded pressures of 40 and 60 mmHg, which was previously proved to be an innocuous stimulus and painful stimulus, respectively (Myers and Greenwood-Van Meerveld, 2012). Each test was repeated three times with a 5-min inter-stimulus interval for recovery. The VMR was quantified by calculating the difference between the area under the curve (AUC) of baseline and the CRD period using Clamp Fit 10.7 (Molecular Devices, United States). AUC was calculated for the 20-s distention period normalized by the 20-s before distention baseline. To compare the change in visceral sensitivity among three different age periods, the relative VMR ratio of each MS group (i.e., the relative AUC/s ratio) was calculated by the mean AUC/s of each MS group normalized by the corresponding NC group (Yi et al., 2017).

Thirty minutes after VMR assessment, the rats were quickly sacrificed and a total of 24 rats’ tissues of the hippocampus were dissected under RNase-free conditions. Total RNA was extracted and then quantified and qualified by Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, United States), NanoDrop (Thermo Fisher Scientific Inc.), and 1% agarose gel. One microgram of RNA of each sample was used to construct sequencing libraries. Afterward, RNA-seq was performed using an Illumina HiSeq instrument according to the manufacturer’s instructions (Illumina, San Diego, CA, United States). In the phase of data analysis, each MS group was compared with the age-matched NC group. Due to the small sample size of each group, the absolute fold change ≥ 1.2 accompanied by a P-value < 0.05 was considered the threshold to detect differentially expressed genes (DEGs). Pathway enrichment analysis was performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG). The DEGs of the three periods of age were put together to draw a Venn diagram to find the specific genes of rats with visceral hypersensitivity from the post-weaning period to adulthood.

The expression levels of mRNA were examined by Quantitative real-time PCR (qRT-PCR) and the primers used are listed in Table 1. Briefly, mRNA was extracted and corresponding cDNA was captured through reverse transcription using PrimeScript™ RT Master Mix (TaKaRa). PCR reactions were carried out with TB Green^® Premix Ex Taq™ (TaKaRa) and performed on QuantStudio7 Flex (Applied Biosystems). The procedure was repeated three times for each specimen. GAPDH was used as the internal reference, and the 2^–ΔΔCt method was used for data analysis.

Protein samples harvested from the hippocampus were lysed by cold lysis buffer (RIPA: PMSF = 100:1) on ice. The concentration of protein was quantified by BCA Protein Assay Kit (Beyotime Biological Co., Ltd.). Twenty micrograms protein samples combined with loading buffer were subjected to SDS-PAGE electrophoresis, transferred onto membranes, and blocked by 5% milk in TBST. Subsequently, the membranes were incubated overnight with primary antibodies, including Netrin-1 (1:500, ab126729, Abcam), DCC (1:1,000, ab273570, Abcam), Neo-1 (1:1,000, 20246-1-AP, Proteintech), GluA1 (1:1,000, ab183797, Abcam), NR1 (1:1,000, ab109182, Abcam), NR2A (1:1,000, ab124913, Abcam), NR2B (1:1,000, ab65783, Abcam), PSD95 (1:1,000, ab36233, Abcam), or Tubulin (1:1,000, ARG62347, Arigo Biolaboratories Co.). On the following day, membranes were incubated with horseradish peroxidase (HRP)-conjugated goat anti-mouse/rabbit IgG antibody (1:5,000, ARG65350/65351, Arigo Biolaboratories Co.) and visualized using an ECL Chemiluminescent Kit (Sigma, United States). Gel image analyses were performed using the ImageJ software.

The lentiviral vectors used in the current study were pCLenti-U6-shRNA (NTN1)-CMV-EGFP(Lenti-shNTN-1) and pCLenti-U6-shRNA(scramble)-CMV-EGFP (Lenti-scramble), which was purchased from Obio Technology (Shanghai).

Since Netrin-1 is enriched in CA1 dendritic spines of the hippocampus, CA1 was chosen as the injection site (Cembrowski et al., 2016). All experiments investigating the efficiency and effectiveness of Lenti-shNTN-1 were conducted on PND35 owing to the most enhanced visceral sensitivity caused by MS during this specific period. On PND21, rats were anesthetized with an intraperitoneal injection of pentobarbital sodium (50 mg/kg), and then the lentivirus vectors (3 ul, 5 × 10⁸ transducing units [TU]/mL) were slowly microinjected bilaterally at the rate of 0.08 μl/min at the following coordinates: anterior–posterior (AP) –3 mm from the bregma, ± 1.8 mm lateral to midline (L), and 3.5 mm ventrally from the skull surface. After injection, the injector cannula was raised by 0.1 mm and held for 5 min to allow for diffusion before extraction. Both groups of MS rats were kept undisturbed for a 2-week recovery and then delivered to behavioral tests and EMG recording (study design shown in Figure 1B). The tissue of the hippocampus or the whole brain was collected 30 min after EMG recording for PCR, Western blot, and fluorescence image analysis. The injection sites and the spread of the viral infection in MS rats were verified at 40 × magnification using an Olympus BX53 light microscope with the 473-nm laser. For the verification of the virus expression, 40 μm thick sections, spanning the entire dorsal hippocampus from Bregma −1.72 to −4.68 mm, were collected according to the stereotaxic atlas of Paxinos and Watson (Paxinos and Watson, 2006).

The data were reported as mean ± SEM. The normal distribution of parameters and homogeneity of variance was evaluated separately by the Shapiro-Wilk test and Levene test. The comparison between the two groups of data was accomplished using the two-tailed Student’s test. Three-group comparisons were carried out using a one-way ANOVA test followed by the least-significant difference test. All statistical analysis was performed using the SPSS 23.0 software (IBM), and p-values lower than 0.05 were considered significant.

Visceral sensitivity was assessed on PND21, PND35, and PND70 (Figure 2), and the typical EMG images were presented (Figure 2E). VMRs of MS rats to both 40 and 60 mmHg CRD stimulation were significantly higher than NC ones in three different age periods, which indicated that MS could induce visceral hypersensitivity in rats from the post-weaning period to adulthood (Figures 2A–C). The relative VMR ratio in MS rats among three age periods was compared. It was found the relative VMR ratio was significantly higher on PND35 than that on PND21 or PND70, indicating that the visceral hypersensitivity induced by MS reached a peak before adulthood (Figure 2D). The anxiety-like behaviors were observed by Open-Field Test (OFT) and EPM (Figure 3). The total traveled distance of MS rats during OFT was shorter than NC rats in all three age periods. Furthermore, MS rats had less entries into the center zone on PND35 and PND70, while no significant difference was shown in the number of rearing (Figures 3A–C). As for the EPM test, MS rats showed less percentage of frequency of entry into open arms and less percentage of time spent in them in all three age periods compared with the NC ones (Figures 3D,E). Moreover, the anxiety index that integrates the EPM behavioral measures also exhibited increased anxiety-like behaviors in MS rats from the post-weaning period to adulthood (Figure 3F). Representative animal track in OFT (Figure 3G and Supplementary Figures 1A–C) and EPM (Figure 3H and Supplementary Figures 1D–F) showed the obvious difference between MS and NC rats.

RNA-Seq was used to explore the specific pathways and genes in the hippocampus involved in the pathological process of visceral hypersensitivity induced by MS in three age periods (Figure 4). KEGG analysis revealed significant enrichment of genes in various pathways (Figures 4A–C and Supplementary Figure 2). Particularly, the axon guidance pathway ranked first on both PND21 and PND35, and also ranked high on PND70. Furthermore, the axon guidance pathway seemed to be the only shared pathway in the top five enriched KEGG pathways of all age periods (Figure 4D). Volcano plots illustrated the distribution and variation of upregulated and downregulated DEGs in different periods (Supplementary Figure 3). To further explore the shared DEGs in different periods, a Venn diagram was drawn, which detected a total of 54 DEGs (Figure 4E). Among the shared DEGs, four axon guidance-related pathway genes were identified, including Netrin-1, DCC, PlexinA4, and MET, which were upregulated in MS rats of different periods (Figures 4F–G). Interestingly, Netrin-1 and DCC were specific ligands or receptors for each other, indicating that their co-upregulation might be a potential mechanism in MS-induced visceral hypersensitivity.

qRT-PCR (Figure 5) and Western blot analysis (Figure 6) were performed to verify the results from RNA-seq at the mRNA and protein levels. Netrin-1 and its receptors DCC, as well as the DCC paralogue Neo-1 and UNC5 families (UNC5A-D), were analyzed. The results showed that both mRNA and protein of Netrin-1 were upregulated from the post-weaning period to adulthood in MS rats (Figures 5A–C, 6). However, the mRNA of DCC and Neo-1 were found upregulated only on PND35, and no significant upregulation of mRNA was seen in UNC5 families. Thus, DCC and NEO-1 were further measured at the protein level in three age periods. The results showed that DCC instead of NEO-1 increased in MS rats on PND35, while none of them altered on PND21 or PND70 when compared with the NC ones. The specific upregulation of DCC provided a profound explanation for the most severe visceral hypersensitivity on PND35.

LTP in the hippocampus has been generally recognized as a trigger in mediating MS-induced visceral hyperalgesia (Chen et al., 2015; Chen et al., 2017), and Netrin-1 has been proved to be a critical factor in enhancing hippocampal LTP (Glasgow et al., 2018). AMPA receptors (AMPARs) and NMDA receptors (NMDARs) are major contributors to the formation of hippocampal LTP (Kullmann and Asztely, 1998; Peters et al., 2021). Therefore, the mRNA and protein expressions of hippocampal AMPARs (mainly comprised of GluA1 and GluA2 subunits; Diering and Huganir, 2018) and NMDARs (including predominant functioning subunits NR1, NR2A, and NR2B; Hansen et al., 2017) in the hippocampus were further evaluated. qRT-PCR results showed that the gene expression of GluA1, NR2A, and NR2B of MS rats were significantly higher than the NC ones from PND21 to PND70, whereas NR1 was upregulated only on PND35 and PND70, and no difference was detected in GluA2 (Figures 5D–F). The results of the Western blot also revealed the increase of GluA1 as well as NR2A and NR2B in MS rats from the post-weaning period to adulthood (Figure 6). In addition, the expression of PSD95, as the marker of the postsynaptic density, was found to increase in MS rats in both mRNA and protein levels compared to NC rats (Figures 5D–F, 6).

To see if Netrin-1 in the hippocampus is essential for the development of MS-induced visceral hyperalgesia, Lenti-shNTN-1 or Lenti-Scramble were microinjected into the bilateral CA1 regions of the hippocampus of MS rats. Due to the limited tolerance of little pups to the surgery, we chose lentivirus intervention on PND21 rather than earlier. And as mentioned above, visceral hypersensitivity reached its peak on PND35, therefore PND35 was the time point to observe related changes after intervention. The results revealed that the expression levels of Netrin-1 mRNA and protein were considerably inhibited on PND35 after hippocampal Lenti-shNTN-1 administration (Figures 7C–E).

The most remarkable finding was that the VMRs to both 40 and 60 mmHg CRD stimulations were deeply compromised on PND35 by hippocampal knockdown of Netrin-1 (Figures 7A,B). The results of behavioral tests also showed the therapeutic effect of Netrin-1 knockdown in aversive emotion. Compared with rats expressing Lenti-Scramble, rats expressing Lenti-shNTN-1 showed more rearing, more entries to center zones, and longer traveled distance in OFT (Figures 8A–C,G). Besides, the percentage of the number of entries into open arms and the percentage of time spent in them were increased while the anxiety index was reduced significantly in EPM after the knockdown of Netrin-1 (Figures 8D–F,H).

To verify the downstream signaling pathway of Netrin-1 in the process of visceral hyperalgesia, the genes and proteins that were previously found to be upregulated accompanied by Netrin-1 in MS rats were examined again (Figures 7C–E). Notably, a significant reduction in the mRNA and protein expression of DCC and GluA1 after administration of Lenti-shNTN-1 was observed. The hippocampal level of PSD95 exhibited a slight decrease after Netrin-1 knockdown. However, Lenti-shNTN-1 failed to influence the expression of NR2A and NR2B. The representative fluorescence image confirmed the successful GFP expression in lentiviruses infected cells in the hippocampus (Figure 7F and Supplementary Figure 4).